L2 Korean Learners’ Socialization into Discourses Around the Non-Honorific ‘Banmal’ Style: Affective and Pedagogical Consequences

Abstract

This study examines L2 Korean learners’ self-reports of their socialization into discourses around the use of two categories of non-honorific (banmal) and honorific (jondaenmal) language. L2 Korean learners (n = 49) of varying proficiency levels completed a questionnaire aimed at capturing their beliefs, attitudes, and practices regarding learning and using banmal. A subset of questionnaire participants (n = 11) were interviewed, and transcripts were analyzed using discourse analysis to understand how banmal is positioned discursively in participants’ self-reported accounts of learning and using L2 Korean. Findings revealed three dominant discourses in learners’ self-reported accounts of their socialization into learning and using banmal: (1) jondaenmal is more important to them than banmal, (2) banmal does not belong in formal learning contexts such as classrooms, and (3) banmal instruction should be delayed until the intermediate or advanced level. Additionally, these discourses were connected to two overarching, at times contradictory, affective responses from participants. While they reported heightened anxiety over when to use banmal, they also described how using it instilled confidence in their sociopragmatic abilities. These findings highlight the connection between the affective experiences of learners and prevailing discourses on particular linguistic forms. Finally, we suggest the need for more integrated approaches to teaching speech styles in L2 Korean classrooms.

1. Introduction

Interest in learning Korean has surged worldwide over the past decade, largely motivated by an interest in Korean popular culture. Although many learners engage in self-study, according to the latest enrollment report published by the Modern Language Association, additional language (L2) Korean enrollments in U.S. higher education increased by 38.3% in the five-year period from 2016 to 2021 (Lusin et al., 2023). This is remarkable, given that, in general, national enrollment numbers for L2 courses in the U.S. declined during this period. In order for learners to (a) access and understand many of the Korean cultural products with which they seek to engage, such as television series, films, and music, and (b) form personal relationships with Korean speakers, learners will need to be able to understand and produce features of spoken conversational Korean. From a pragmatics standpoint, this includes the use of non-honorific speech, colloquially referred to as, ‘banmal’ or, literally, “half-speech.” Conventionally, banmal has not been a focus of Korean foreign language (KFL) education; instead, classes and materials have favored the honorific category of speech known as jondaenmal (Brown, 2010; Choo, 1999; Kiaer et al., 2022). The current study aims to explore learner reports of socialization into the discourses around L2 Korean honorifics and some of the self-reported affective impacts of these discourses on their orientation toward their studies.

In addition to new linguistic forms, acquiring an additional language involves a process of socialization into the cultural norms, language ideologies, and expectations of the target culture. Language socialization has been defined as “the process by which novices or newcomers in a community or culture gain communicative competence, membership, and legitimacy in the group” (Duff, 2007, p. 310). This process occurs as novices or language learners interact with more expert members of a particular community, such as first-language (L1) speakers, instructors, or more advanced L2 speakers. By doing so, the process of language socialization “is deeply affected by the process of becoming a competent member of society” (Schieffelin & Ochs, 1986, p. 168).

For L2 learners, this process can be complicated by a number of factors, including various aspects of their identities, the social positions available to them in the target culture, and the prevalent discourses in the target society regarding what constitutes appropriate language for L2 speakers. In languages such as Korean, with highly developed honorific systems that are intrinsically connected to speaker identity and social position, the language socialization of newcomers can be complex. Brown (2010) notes that “…knowing when to use honorific forms is vital not only for a Korean speaker to present him/herself as someone who is proficient in the language, but also to claim an identity as a competent member of society” (p. 36). From a language socialization perspective, becoming a competent Korean speaker involves acquiring honorifics through interactions with L1 speakers and with Korean language instructors. Despite their importance, L2 Korean speakers’ access to both highly honorific and non-honorific forms has been shown to be restricted in their classrooms, learning materials, and in interactions with L1 Korean speakers (Brown, 2010; Choo, 1999; Kiaer et al., 2022).

This restriction seems to be related to discourses prevalent within Korean foreign language (KFL) education and Korean society about what constitutes appropriate language use for L2 speakers. Brown (2010, 2016) and Choo (1999) analyzed Korean language textbooks and found that non-honorific speech styles were underrepresented compared to honorific speech styles. Brown (2010) similarly found that the teaching of non-honorific speech styles was delayed until the intermediate level in the three textbooks he reviewed, and that there were marked differences in how Korean vs. non-Korean characters use honorifics in the textbook dialogues.

This restriction of honorifics has been shown not only in textbooks but also, to some degree, in interactions with L1 speakers. In a study abroad context, the L2 Korean learner participants in Brown’s (2011) study reported that Koreans used honorific and non-honorific language differently with them than with fellow Koreans, which the participants attributed to their identities as foreigners in Korean society. Kiaer et al. (2022) interviewed L1 Korean speakers and found that they reported a belief that certain aspects of the honorific system, specifically kinship titles, should be used differently with foreigners than with L1 Korean speakers. Byon (2003, 2006) analyzed classroom discourse and found that it was a powerful place for L2 and heritage students to be socialized into Korean hierarchical cultural values through the use of certain features such as speech styles and first-person pronouns. Building on this research, the current study analyzes self-reported experiences of both classroom-based and independent, self-directed L2 Korean learners, focusing on how they are socialized into discourses surrounding the learning and use of non-honorific language.

2. Literature Review

2.1. Second Language Socialization

Language socialization is rooted in the idea that learning occurs through interaction between novices and experts. Duff (2007) notes that “research in language socialization generally recognizes that language and literacy learning involves explicit or implicit socialization through linguistic and social interaction into relevant local communicative practices or ways of using language and into membership in particular cultures or communities, with their own values, ideologies, and activities” (p. 310). For L2 learners, language socialization is a process of both acquiring linguistic forms with which to speak the target language and also, through interaction with L1 speakers and instructors, learning how to be a “competent member of society” (Schieffelin & Ochs, 1986, p. 168). This makes honorific and non-honorific language a particularly apt area of study for language socialization, as honorifics are intimately tied up with knowing one’s role in an interaction and then acting and speaking in a way that is appropriate for that role. In this sense, honorifics are both the means by which L2 Korean speakers are socialized into certain cultural norms and the prerequisite for legitimate participation in various L1 communities.

Previous research has shown that the L2 classroom is a domain in which Korean learners are socialized into using honorific language and hierarchical cultural norms. For example, Byon (2003) found that the use of the polite speech style ending yo was a powerful tool for socialization in Korean heritage speaker classes. Additionally, Byon (2006) showed that teachers socialized L2 Korean students into hierarchical Korean sociocultural values through several aspects of the honorific system, such as honorific verb infixes, personal pronouns, and assertive directives, in both explicit and implicit ways. He notes that “it is reasonable then, to assume that teacher talk has a crucial role in socializing KFL students into Korean sociocultural values” (p. 284). Outside the classroom, Suh (2020) found that honorifics were used in Korean-American families to socialize heritage speakers of Korean by encouraging compliance, demonstrating culturally appropriate ways to give directives, and providing practice in using those culturally appropriate directives. Her study also showed that family members used honorifics to foster and support development of the Korean.

While L1 norms are a large part of language socialization, the discourses that exist in the target society around L2 speakers may also result in socialization into L2 speaker-specific discourses. In other words, L2 speakers may be socialized into separate discourses that exist specifically about L2 language use and learning, and which differ from those surrounding L1 language use. L2 speakers may find themselves caught between discourses around expected L1 language use and expected L2 language use. For example, Maa and Burns (2021) found that their case study participant, an L2 learner of Japanese, struggled to reconcile two opposing discourses in her learner socialization: (1) that learners should aspire to use Japanese like native speakers; and (2) that Western learners, in particular, should stick to a specific, codified “foreigner” speech style. It is important to investigate how learners are socialized into these sometimes conflicting expectations around their language use, as well as how their learning contexts may influence their beliefs about what constitutes appropriate honorific and non-honorific language.

2.2. L2 Korean Honorific and Non-Honorific Language

Honorifics are linguistic features that allow speakers to signal their relative position to either their interlocutor or the referent of an utterance, using the morphology, lexicon, and prosody of the language. The Korean honorific system is made up of many parts, which include honorific address titles, honorific verb infixes, humble pronouns, speech styles, and honorific lexical items. Therefore, there are often many ways to craft a grammatically correct sentence that may nevertheless be sociopragmatically inappropriate given the situation. Choices about which parts of this system to use (or not) are based on social and contextual considerations such as age, gender, social position, formality of a situation, and psychological closeness. These choices are highly cultural in nature and can be challenging for L2 Korean speakers to make, given the linguistic complexity involved and their sometimes vague social position in Korean society.

Among the many linguistic features that make up the honorific system in Korean, both L1 and L2 speakers typically practice the “Korean folk notion of dividing speech into two rough categories: jondaemal (‘respect speech’) and banmal (‘half-speech’)” (Brown, 2013, pp. 273–274). Most KFL curricula and pedagogical materials focus on teaching the category of polite, honorific speech known as jondaenmal at the expense of the non-honorific speech category called banmal. Jonadenmal “is characterized by the use of honorific speech styles, the strict application of referent honorifics, name/pronoun avoidance, wordiness, and regular prosody” (Brown, 2011, p. 274). On the other hand, banmal “is characterized by the use of non-honorific speech styles, less strict application of referent honorifics, use of personal names, vocative forms and pronouns, reduced wordiness and less regular prosody” (Brown, 2011, p. 276). While certain relationships may imply the use of banmal from the start, such as between students of the same grade level or two people of the same age, the appropriateness of its use can also evolve over time as two people become closer if the context allows it. In these instances, one party may explicitly ask if it would be okay to “lower their language” or “speak comfortably” to gain the consent of their interlocutor. Choo (1999) suggests that the continued use of jondaenmal with a close friend “can hurt the feelings of those who wish to be addressed in a more friendly way” (p. 79). For L2 Korean learners, this can be a difficult and sometimes uncomfortable choice. In Brown’s (2011) study of Korean language learners studying abroad, he noted that “L2 speakers may be over-conservative in setting the parameters as to when it is appropriate to move from a register or strategy that is perceived as “polite” to one that is seen as less polite or more informal, casual, etc.” (p. 95). This may impede their ability to form new personal relationships or to deepen existing ones.

Following Brown (2013), we use the term banmal to refer to the category of informal, intimate, and non-honorific language that is characterized by the absence of many deferential features of honorific speech and is generally used in informal settings and close relationships. While recognizing that the boundaries between banmal and jondaenmal can be fluid and that speakers often mix features of the two categories, we elect to use the terms banmal and jondaenmal as they are the ones that learners are most likely to hear in conversations with L2 speakers. Given that this study focuses on the discourses surrounding the learning and use of banmal, we describe the features of this speech style in detail below.

In banmal, the two primary speech styles are hayla-chey and hay-chey, which differ morphologically based on the conjugation of the sentence-final verb and sociopragmatically based on the speaker’s relationship with the listener and their social context. The first, hayla-chey, is considered the least deferential style. It can also be used as “a generic style used in writing to a general non-specified audience” (Brown, 2011, p. 25) as well as when “speakers convey newly perceived or retrieved information” and with “expression of explanation” (Lee, 1991; Choo, 2006; in Brown, 2011, p. 26). Even between people who use banmal, the use of this style, particularly its interrogative form, can be somewhat restricted due to the fact that it is often used when the speaker orients negatively to the information being conveyed (Choo, 2006; Lee, 1991).

Table 1. Overview of speech styles in modern, spoken Korean.

	Speech Style	Example Sentence “This Is Coffee.”
반말 Banmal	하라체 Hala-chey	이것은 커피다 igeoseun keopida
	해체 Hay-chey	이것은 커피야 igeoseun keopiya
존댓말 Jondaenmal	해요체 Hayo-chey	이것은 커피에요 igeoseun keopieyo
	합쇼체 Hapsyo-chey	이것은 커피입니다 igeoseun keopiimnida

shows the breakdown of the most commonly used speech styles in banmal and jondaenmal, along with sample sentences.

The second banmal speech style, hay-chey, is used in conjunction with hayla-chey. Brown (2011) notes that this “has become the most commonly used non-honorific speech style” (p. 26) in modern Korean. Its conjugation involves removing the final “yo” syllable from the hayo-chey speech style in jodaenmal, so it is relatively easy to conjugate and is the most noticeable marker of whether one is speaking banmal or jondaenmal (Brown, 2011). In addition to the hayla-chey and hay-chey speech styles, there are several other features that make up banmal. The first is the use of first names or first names with the vocative particle -a/ya. When speaking in jondaemal (formal, honorific speech), the use of first names is highly restricted, but in banmal, it can be appropriate to use first names as well as the first-person pronoun na as opposed to the humble first-person pronoun ce. In addition to the first-person pronouns, the use of the second-person pronoun ne can also be appropriate in banmal, whereas it is virtually, if not entirely, restricted in jondaemal.

In L2 Korean classrooms, learners are largely taught the hayo-chey speech style using jodaenmal. However, it is worth noting again that the only morphological difference between the more polite hayo-chey and the hay-chey speech styles of banmal is dropping the final “yo” syllable. The ease with which hay-chey can be conjugated serves as a counterargument to the prevailing discourse that L2 Korean curricula should exclusively focus on jondaenmal in order to simplify the language for learners. Brown (2010) notes that many of the other grammar points introduced to students at the beginner level are much more complex than the conjugation of the hay-chey speech style. In addition to speech styles, studies such as Chen and Brown (2022) have shown that learners can achieve a relatively high degree of success in understanding the form-context mapping of humble vs. plain first-person pronouns, although they lacked “native like” understandings of their indexical meanings. This suggests that learning various aspects of the honorific system is well within learners’ capabilities, and that it is also crucial for developing their sociopragmatic understanding of these forms.

2.3. The Present Study

These studies provide a foundation for understanding the backdrop of beliefs against which L2 speakers must negotiate an understanding of this important pragmatic system. However, while previous studies have taken various approaches to investigating L2 honorific use and socialization of certain honorific features, they have not necessarily addressed how L2 Korean learners characterize the experience of being socialized into discourses around learning and using honorifics. By taking a learner-centered approach, we also seek to contribute a new perspective on language socialization by investigating learners’ affective reactions to the discursive socialization they report experiencing in their various learning contexts. While foreign language anxiety (Horwitz et al., 1986) has been widely researched, other affective aspects of the foreign language learning process have not been addressed to the same extent. Dewaele and Li (2020) note that “scant attention had been paid to emotions in second language acquisition” (p. 1). The Douglas Fir Group (2016) emphasizes that “language learning is an emotionally driven process at multiple levels of experience” and that these emotions are “enmeshed with identity, agency, and power, all central in the learning and teaching of languages in today’s multilingual world” (p. 36). Considering this, it is crucial to investigate not only what discourses learners report experiencing as they learn L2 Korean honorifics, but also how these beliefs may influence their self-reported affective experiences of L2 learning. To this end, this study sought to answer the following questions.

RQ1: 

What discourses are present in participants’ self-reports of being socialized into learning banmal as L2 Korean learners?

RQ2: 

What effect does the discursive positioning of banmal in participants’ accounts of L2 Korean learning have on their self-reported affective experiences?

3. Methodology

3.1. Participants

Participants for this study were recruited through online Korean language learning communities on Facebook and Instagram, as well as through personal and professional contacts via email. The inclusion criteria were as follows: Korean language learners of any level, over the age of 18, located in the US, and fluent in English. We aimed to include participants who had learned in both traditional classroom settings and through independent, self-directed study. The questionnaire received 49 responses from those eligible to participate, and 14 participants were selected for follow-up interviews. However, three participants chose not to comply with the study requirement of allowing video recording; therefore, 11 interviews were included in the analysis presented here. Given the diverse learning contexts of the participants and the study’s focus on participants’ own perceptions of their learning experience, self-rated proficiency was chosen as an appropriate way to understand their perceptions of their own language ability. Among the 49 survey participants, 19 rated their proficiency as minimal, 16 as elementary, 7 as intermediate, and 6 as advanced. An overview of the interview participants is presented in

Table 2. Interview Participants.

Name	Gender	Age	Race	L1	Length of Study in Years	Self-Rated Korean Proficiency
Arna	Female	23	Indian	Kannada, Hindi	3	Elementary
Becky	Female	63	White	English	2	Elementary
Carrie	Female	32	White	English	Less than 1	Elementary
Charlie	Male	34	White	English	4	Elementary
Ila	Female	20	Asian/Indian	English	3	Intermediate
May	Female	46	White	English	20	Intermediate
Minji	Female	20	Asian	English, Korean	1	Advanced
Rachel	Female	26	White	English	9	Intermediate
Sarah	Female	28	White	English	5	Advanced
Veronica	Female	52	Hispanic	English, Spanish	7	Elementary
Vy	Female	23	Asian	English, Vietnamese	5	Elementary

. All names are pseudonyms.

3.2. Data Collection Methods

This mixed-methods study utilized an online questionnaire (49 respondents) consisting of background information and questions about participants’ beliefs towards banmal, as well as follow-up semi-structured interviews with 11 participants. A mixed methods approach was chosen as a large questionnaire allows us to obtain a broader picture of learners’ experiences and beliefs about learning and using banmal. Follow-up qualitative interviews with 11 participants provide access to deeper and more nuanced descriptions of participants’ experiences with banmal.

3.2.1. Questionnaire Instrument

The questionnaire consisted of three sections: first, participants answered 15 background questions related to their demographic information and experiences with language learning. The second section included seven questions about the degree to which they had used certain resources in their learning of banmal. The third section consisted of 25 Likert scale items that asked participants to rate, on a 5-point scale, the degree to which they agreed with statements about learning and using banmal. For example, “I believe that it is important for Korean language instructors to teach banmal.” At the end, they were asked to leave their email addresses if they were interested in participating in a follow-up interview. The questionnaire can be found in Appendix A.

3.2.2. Qualitative Interview

The first author conducted semi-structured interviews in order to gain deeper insights into the participants’ beliefs and questionnaire responses. When possible, she asked follow-up questions or attempted to solicit narratives. The interviews included questions such as “How has your understanding of banmal changed over time?”, and “Do you think it is important for Korean learners to learn banmal early on? Why or why not?”. The interviews were conducted via Zoom and included both audio and video recordings. They lasted 15 min to one hour. This study focuses on the results of qualitative interviews, with supporting evidence from quantitative questionnaire data. The interview protocol can be found in Appendix B.

3.3. Data Analysis Procedures

The quantitative data were analyzed in R-Studio (R version 4.3.1), and descriptive statistics are presented together with the qualitative data to provide a more holistic picture of the analysis. The qualitative interviews were analyzed using discourse analysis (Waring, 2018; Wodak & Meyer, 2015). This approach was chosen as the most appropriate way to go beyond what participants say at the sentence level and discover larger phenomena that exist at the discourse level. Following Saldaña (2021), to address each of the research questions, the interviews were fully reviewed several times, first noting each time the participant described some belief held, either by them or reported to be held by their instructors or learning materials, about speech style, banmal, or jondaenmal. These were then categorized thematically, and a set of codes was created, which were used to code the interview data in a second round of coding. Three overarching discourses were identified regarding the participants’ socialization process into learning banmal, which existed across the participants, their instructors, and their learning materials. These discourses are presented in detail below. The final round of coding focused on the learners’ affective experiences involved in learning or using banmal, which resulted in two main thematic categories, which are presented in detail below.

4. Results

4.1. Discourses Around Learning Banmal

Discourse analysis of the interviews and corresponding questionnaire responses revealed three main discourses surrounding L2 Korean speakers’ learning and use of jondaenmal and banmal: (1) Jondaenmal is more important for L2 Korean speakers than banmal; (2) banmal does not belong in formal learning settings; and (3) banmal instruction should be delayed to the advanced or intermediate level. In the following sections, we detail each of these discourses through representative quotations from the participants, as well as corresponding data from the questionnaire.

4.1.1. Jondaenmal Is More Important for L2 Korean Speakers than Banmal

The first discourse identified in the data was that jondaenmal is more important for L2 Korean speakers than banmal. In the interviews, the participants reported being socialized into this discourse both explicitly by their teachers and tutors and implicitly through its almost complete absence in their learning materials. As shown in

Table 3. Learners’ beliefs about relative importance of banmal instruction.

Survey Item		Mean	SD
6	I believe that it is more important for Korean learners to learn banmal than jondaenmal.	2.7	1.3
8	I believe that it is more important for Korean learners to learn jondaenmal than banmal.	3.8	1.0

, the participants slightly disagreed that banmal is more important for learners than jondaenmal (m = 2.7) but agreed that jondaenmal is more important for learners than banmal (m = 3.8).

Regarding their explicit socialization into this discourse, some interviewees reported being directly told by their teachers that jondaenmal was more important for them to learn. Charlie and Carrie, who both have experience in formal Korean classes or with individual tutors, reported being cautioned against using banmal in favor of jondaenmal. Charlie, an avid language learner, referred to this preference for jondaenmal as the “golden rule” and reported hearing this explicitly from a variety of sources. He said, “the golden rule in all the books and all the Korean teachers and YouTube videos is that it’s better to be too formal and sound a little goofy than it is to be, you know, informal.” Carrie’s descriptions of her explicit socialization into this discourse were particularly powerful. For instance, she noted that she intentionally avoids using banmal even when its use may be appropriate, citing clear directives from her instructor to do so. While first describing how her teacher is “very big on” teaching her “the different levels,” she reported that he also explicitly told her “you shouldn’t use banmal,” “don’t ever use it,” “you don’t use this,” “you don’t have to do it.” In fact, Carrie stated six times throughout her interview that her teacher had told her not to use (or that she did not have to use) banmal at all. Given the high prevalence of banmal in L1 discourse and its role in creating close, equal, and intimate relationships, it is noteworthy that Carrie was explicitly instructed not to use it.

Not only were the participants explicitly socialized into this discourse by their teachers and tutors, but they were also implicitly socialized into the idea that jondaenmal is more important for them through their learning materials. Vy, Veronica, Ila, and Becky reported that in-depth explanations of banmal were scarce in their textbooks and learning materials. For example, Vy, who took Korean classes in college, reported that because the textbooks only taught jondaenmal, she thought that yo and supnita [speech style endings in jondaenmal] were the only ones that existed. Veronica and Ila both stated that while the textbooks start off by acknowledging that there is a distinction between honorific and non-honorific forms, as they only “teach you in the formal way” (Veronica) and that “most of the examples and the way that they taught you grammar was for formal speech” (Ila). Becky, an older learner who studies with a tutor, described the position of banmal in her learning materials as “pushed aside” and asked “do grammar books even cover that? The books I’ve come across don’t.” Thus, as instructors and learning materials acknowledge banmal’s existence but marginalize its instruction, learner perceptions indicate that they are implicitly socialized into the discourse that jondaenmal is simply more important than banmal.

4.1.2. Discourse #2: Banmal Does Not Belong in Formal Learning Settings

The next overarching discourse identified in the participants’ discussion was that banmal does not belong in formal learning settings. Their reports indicate they were largely socialized into this discourse implicitly through the contexts in which they could and could not gain access to banmal. Specifically, most of their exposure to banmal came through casual social encounters, media, and non-traditional learning contexts like Korean dramas, TikTok, and Instagram. On the quantitative survey items that asked participants to rate the degree to which they used a particular resource when learning banmal, they reported that they used media and speaking with friends more than resources in a classroom, from a textbook, or from a tutor. Figure 1 shows the distribution of participant responses. Roughly 65% of participants reported that they only ‘minimally’ or did ‘not at all’ learn banmal through textbooks, roughly 63% either ‘minimally’ or did ‘not at all’ learn banmal from a teacher, and 73% reported that they either ‘minimally’ or did ‘not at all’ learn banmal in a classroom. In contrast, participants reported learning banmal through media or speaking with friends either ‘somewhat’ or ‘a lot’ at a higher rate. Roughly 71% reported using media either ‘somewhat’ or ‘a lot’ to learn banmal. Speaking with friends had a more even distribution, with 51% reporting using this method either ‘minimally’ or ‘not at all,’ whereas 48% used speaking with friends either ‘somewhat’ or ‘a lot’ to learn banmal. This aligns with previous textbook analyses (Brown, 2010, 2016; Choo, 1999), which have shown that banmal is underrepresented in Korean language textbooks.

While the participants overall reported learning banmal ‘not at all’ or ‘minimally’ in a classroom or in a textbook, the Likert scale survey items which targeted learner beliefs about the importance of banmal instruction show that participants weakly agreed that L2 Korean speakers should be taught banmal by teachers and in their textbooks. A summary of these responses is presented in

Table 4. Learners’ beliefs about banmal instruction.

Survey Item		Mean	SD
2	I believe that it is important for Korean language instructors to teach banmal.	3.6	1.0
3	I believe that it is important for Korean language textbooks to teach banmal.	3.7	1.2

. These results show that, despite the belief that it is important for L2 Korean language instructors and textbooks to provide instruction on banmal, the reality for these Korean learners is that they did not receive banmal instruction in classroom settings but rather through alternative methods.

In the interviews, Ila, Arna, Carrie, and Vy described having to turn to supplementary sources to access banmal. One important tool that participants used to learn banmal was Korean television dramas. Ila described how she began to learn the difference between honorific and non-honorific speech by noticing how the characters style-shifted in dramas. She stated, “in TV shows, you would notice that they don’t say yo and stuff, and I kinda understood like vaguely when people use like banmal versus like jondaenmal.” Arna similarly pointed out that, in contrast to textbook examples that are “not real-world applicable,” “the banmal that I see people use in like K-dramas seems more right to me and that’s what I try to use.” Carrie also described how her “education” on banmal came through her own research after watching Korean dramas. “It all kind of started from hearing in K-dramas like ‘let’s just speak casually’ I was like, what does that mean, and so I kind of like looked into it on my own.” In addition to Korean dramas, non-traditional learning domains and social media sites such as Instagram and TikTok also served as contexts in which participants gained access to banmal instruction. Vy noted that while she has not seen banmal being taught in textbooks, online resources such as the popular website Talk To Me in Korean, as well as TikTok and Instagram, “show all those levels.”

4.1.3. Discourse #3: Banmal Instruction Should Be Delayed to the Intermediate or Advanced Level

The final prevalent discourse in learner reports of their socialization into honorifics was that banmal instruction should be delayed until the intermediate or advanced level. Participants conceded, at times, that the concept of banmal could be introduced early on, but their responses also indicated that they believed beginners did not need to know how to conjugate it. This seems to be related to the belief that banmal is easy to learn, can be picked up at any time, and does not require a lot of specific attention.

Table 5. Learners’ beliefs about the order of banmal instruction.

Survey Item		Mean	SD
5	I believe that Korean learners should learn banmal first before learning jondaenmal.	2.8	1.4
7	I believe that Korean learners should learn jondaenmal first before learning banmal.	3.9	1.1

shows the survey responses for the items that asked whether jondaenmal or banmal should be learned first. Item 5 shows that participants disagreed (m = 2.8) with the statement that banmal should be learned first, whereas they agreed (m = 3.9) with the statement that jondaenmal should be learned first.

Interview participant Rachel described being explicitly socialized into this discourse by her Korean language teacher in college. She stated that her teacher “mentioned banmal to us and she maybe did like a small unit on it, but her belief, and I agree, is that it’s better to learn kind of the longer, more formal version first and then she said you can always just shorten it later.” While it is true that conjugating sentence-final verbs in banmal simply requires dropping the final yo syllable, there are many other aspects of banmal use that are much more complicated, including knowing when and with whom it is acceptable to use it, how to negotiate its use, and its other grammatical features. Becky similarly stated that learning banmal “could be helpful as you’re getting more to like intermediate or like advanced level, right like once you’ve been learning Korean for a while or you’re showing or demonstrating a good amount of proficiency.”

Sarah, the most advanced non-heritage speaker participant, stated “after you understand that yo form and that and the differences in the uh in the hierarchy and in politeness levels um then I feel like banmal can be introduced.” She connected this to an increase in expectations for more proficient L2 speakers to be able to use banmal and jondaenmal as Koreans do. She stated, “once you reach the upper intermediate to advance to fluent they expect you to speak the way that they speak.” Carrie also stated, “I think that it’s important to learn the concept of banmal early on, I don’t think it’s necessarily important to learn how to conjugate things using banmal or anything like that.”

One noteworthy finding was that participants’ reports of socialization into this discourse seemed to be somewhat mediated by the learning context. Among the interview participants, those who had experience learning in a formal setting, either in a classroom or with a tutor, were more likely to report that banmal should be introduced later on in the learning process, whereas those who had mostly self-studied or learned in heritage contexts were more likely to report believing that banmal should be introduced early on. With the exception of May, this was true for all the interview participants. A summary of the participants’ learning contexts and beliefs about when to introduce banmal is presented in

Table 6. Learning context and beliefs about when to learn banmal.

Name	Learning Context	Banmal Should Be Learned
Rachel	Classroom	Later
Sarah	Classroom	Later
Vy	Classroom	Later
Becky	Korean Tutor	Later
Carrie	Korean Tutor	Later
Charlie	Korean Tutor	Later
Minji	Heritage Speaker	Early
Arna	Self-study	Early
Ila	Self-study	Early
May	Self-study	Later
Veronica	Self-study	NA

.

This effect of learning context on beliefs about when banmal should be introduced suggests that learner reports indicate being socialized not just into the use of certain linguistic forms, but also into the prevailing discourses related to language learning. While participants overwhelmingly reported being socialized, implicitly or explicitly, into the discourse that jondaenmal is more important than banmal, not all of them accepted this passively. For example, Ila reported wishing that she had had access to materials that integrated banmal early on. She stated, “I wish that resources online that taught formally sort of integrated it very early on.” Arna, an avid Korean drama fan and self-studier, described her belief that introducing banmal early on can be used as a tool to help beginning learners better understand Korean because banmal’s conjugations are generally simpler than those of jondaenmal.

Therefore, while L2 Korean speakers may be socialized into the existing discourse within KFL education that jondaenmal is more important than banmal, they do not always unilaterally accept this. Instead, it seems that those who report being explicitly socialized into this discourse by their teachers and classrooms are more likely to also believe that banmal instruction should be delayed, whereas those who self-study, and may only be implicitly socialized into this discourse by their learning materials, believe that banmal should be learned early on.

4.2. Affective Consequences of Discourse About Banmal in KFL Education

The analysis thus far has shown that participants accounts indicate explicitl and implicit socialization into the discourses that (1) jondaenmal is more important for them than banmal, (2) banmal does not belong in formal learning contexts such as classrooms, and (3) banmal instruction should be delayed to the intermediate or advanced level. In order to go a step further and investigate how these discourses may influence the learning experience, the interviews were also analyzed for participants’ affective responses. Our analysis of the questionnaire and interview data revealed two main findings. First, the participants reported high levels of anxiety regarding the use of banmal. Second, participants described feeling a greater sense of confidence and inclusion when they were able to use banmal successfully.

4.2.1. Anxiety

Charlie, Rachel, and Carrie all reported some level of anxiety or nervousness associated with using banmal. Charlie described how, when studying a foreign language, he is always excited to start using informal language forms because he sees that as a way to express his own identity as a friendly, casual American. However, he stated, “In Korean, I have this anxiety about it like visceral anxiety cuz it could be extremely offensive if you do it.” Later on, he explained that part of the reason for this anxiety is that while Koreans have a “natural” sense of when they can ask to switch to banmal, he gets anxious about the idea of asking someone to use it with him. “Amongst each other, it’s more natural, they know when to ask and when they can switch, and I just have a visceral anxiety about ever trying to ask anyone to use it.” This anxiety went so far as to prevent him from using it even in his own head. He noted, “I try, even when I’m like practicing talking in my head […] I will still use the yo form because I don’t want to accidentally slip out banmal if I’m talking to somebody.” It’s because of this anxiety that he reports only using banmal with his dog, who will not critique him or disclose errors to anyone, remarking that “he [the dog]’s not going to tell anybody.”

Carrie used particularly strong language regarding the anxiety of using banmal and the risk of potentially offending someone. Given that she was explicitly told by her teacher that jondaenmal was more important and that she should not use banmal, her experience of socialization into these discourses may contribute to this fear. She stated, “I’m just so scared of offending people I don’t-- that’s like-- my biggest fear is like accidentally offending somebody um when you know that’s not my intention ever.” She also described how her Korean tutor “terrified” her regarding the use of banmal, saying “[he] put the fear of god into me.” Nevertheless, she believes his approach was appropriate as it taught her “not to be cavalier about it.” Although Carrie’s instructor may have been well-intentioned, his positioning of banmal as inappropriate for students has powerful affective consequences for Carrie, which, in turn, may affect her development of pragmatic and interactional competence as well as her ability to form close relationships with Korean speakers.

Two participants mentioned age-related anxiety regarding their speech style in Korean. Rachel described a situation from her study abroad homestay in Korea. She stated that she would often have a hard time at the dinner table switching between jondaenmal when talking to her host parents and banmal when addressing her host siblings, who were much younger than she. Rachel said, “I would often just use the yo ending automatically, even with the little kids, which was totally weird and wrong, but I was more afraid of the reverse happening because then that would be disrespectful.” Similarly, Becky, an older learner at age 63, noted a connection between her age and some nervousness about initiating the use of banmal even with other Korean learners. She said, “I don’t know what they would think if they if I tried to do that with them, they’d be like oh you trying to be my friend you old lady.” Rachel and Becky illustrate how speech style plays a key role in how learners must navigate the tension between wanting to remain respectful while also desiring to form close relationships.

This discomfort around knowing when to initiate the use of banmal was also shown in the questionnaire data (See

Table 7. Confidence using banmal.

Survey Item		Mean	SD
13	I feel confident knowing when I can use banmal with someone.	3.33	1.3
15	I feel comfortable initiating the use of banmal with a fellow Korean language learner.	3.45	1.3
17	I feel comfortable initiating the use of banmal with a native Korean speaker.	2.9	1.4

). Participants reported weak agreement with the statement “I feel confident knowing when I can use banmal with someone” (m = 3.33). Interestingly, they reported higher levels of comfort initiating the use of banmal with a fellow Korean language learner (m = 3.45) compared to initiating the use of banmal with a native Korean speaker (m = 2.9). When considering these responses together with the interview data, it seems that some discomfort around initiating the use of banmal with an L1 Korean speaker may come from a fear of inadvertently giving offense.

4.2.2. Feelings of Confidence, Fluency

Both the questionnaire responses and interview data show a connection between participants’ beliefs about their ability to use banmal and their feelings of being fluent speakers of Korean. An overview of the questionnaire responses is presented in

Table 8. Learner beliefs about banmal’s relationship with fluency.

Survey Item		Mean	SD
18	I believe that I need to know how to speak banmal well in order to feel that I am a fluent Korean speaker.	4.0	1.0
19	Being able to use banmal well would make me feel like a fluent Korean speaker.	3.8	1.1
20	A Korean language learner needs to be able to use banmal well in order to be considered a fluent speaker.	3.8	1.2
21	When I meet someone who can use banmal well, I feel that they are a fluent Korean speaker.	3.6	1.1

.

As shown in the table, the participants most strongly agreed with the statement “I believe that I need to know how to speak banmal well in order to feel that I am a fluent Korean speaker” (m = 4.0). Of all the Likert scale items, this one had the strongest agreement, which provides an important contrast to the discourses reported above, which minimize banmal’s importance for learners.

This connection between banmal and feelings of being a fluent or confident Korean speaker was also evident in the interview data. Carrie, Ila, and Arna described how the act of using banmal made them feel like confident speakers. Arna noted that “in banmal I feel a lot more confident in myself.” Similarly, Carrie mentioned that, in addition to making her a more self-assured speaker, using banmal also made her more confident in her relationships with her interlocutors. Ila reported that she felt she needed to be able to use banmal and jondaenmal in order “to reach full proficiency.” She went on to describe how using banmal would allow her to move beyond simply a Korean learner identity and embody a Korean-speaker identity. Renfroe (2024) similarly showed that banmal allows L2 Korean speakers to authentically express various aspects of their identities. In this study, the participants’ feelings of fluency and confidence when using banmal appear to be connected to their ability to authentically represent themselves in Korean.

5. Discussion

In this paper, we argue that socialization takes place not only at the linguistic level but also at the discourse level. Through interactions with experts such as instructors, tutors, and L1 speakers, L2 Korean speakers report being socialized into prevailing discourses about learning and speaking banmal and jondaenmal. According to participant accounts, this socialization takes place both explicitly, with learners stating that their teachers tell them not to use banmal, and implicitly, through its marginalization in textbooks and instruction. As shown through this analysis, participants report affective reactions as a significant component of the process of socialization into these discourses. Whether intentional or not, the discourses of jondaenmal’s greater importance appear to discourage learners from using banmal even when it would be appropriate and, in the case of some learners, cause “visceral” anxiety at the mere thought of using it. At the same time, when they successfully employ banmal, participants report a sense of feeling included or of belonging to the category of “proficient Korean user” due to the discourse that banmal should be delayed or reserved for only L1 or advanced speakers of L2 Korean.

By systematically documenting these discourses, we do not intend to suggest that they are objectively flawed or imply that Korean language instructors and textbooks intentionally restrict L2 learners’ access to banmal for ulterior motives. In the case of the discourse that jondaenmal is more important for L2 speakers than banmal, we acknowledge that there are certainly many reasons why instructors may choose to prioritize teaching honorific forms before non-honorific ones. However, as shown through the participants’ affective responses to these discourses, the near-total exclusion of banmal from the beginner levels of instruction and its marginalization at higher levels may have the unintended consequence of scaring learners away from using a speech category that is highly prevalent in L1 discourse. In fact, not only is banmal necessary for learners due to its high frequency in L1 discourse, but also because it reflects the fact that “in addition to the concepts of ‘respect’ and ‘politeness,’ the concept of ‘closeness’ is also highly valued in Korean culture” (Choo, 1999, p. 79).

Curriculum designers, textbook authors, and instructors necessarily make choices about which aspects of language to teach and in what sequence, due to a variety of real-world constraints. They do so, at least in part, based on their conscious or unconscious beliefs regarding which aspects of the language are most appropriate for students. This phenomenon is not unique to Korean language education. Burns (2018), for example, found that U.S. varieties of Spanish are both implicitly and explicitly de-legitimized in L2 Spanish classrooms and textbooks and argued that standard language ideologies underpinned that discursive marginalization.

In the case of L2 Korean honorifics, our study aligns with previous research showing that the pedagogical decision to minimize banmal, especially at beginner levels, often has an ideological component. For example, Brown (2010) found that the presentation of honorific and non-honorific language in L2 Korean textbooks reflects something closer to foreigner talk than authentic L1 Korean. He notes that “it seems clear that these ideologies and preconceptions have influenced the decision to present honorifics in a way that cannot fully represent Korean-style patterns of interaction” (p. 48). He found that all three textbooks he analyzed delayed the teaching of non-honorific speech styles to at least the intermediate level, and that the dialogues which did depict non-honorific speech style use tended to heavily feature Korean characters rather than the foreign characters, suggesting that non-honorific speech is not expected from L2 Korean users. While previous studies (Brown, 2010, 2016; Choo, 1999; Kiaer et al., 2022) have outlined how ideologies about what constitutes appropriate language for L2 Korean speakers influence the kind of language they have access to in their textbooks and even in interactions with L1 speakers, this study contributes a new perspective by considering participants’ accounts of their socialization into discourses around honorifics, as well as their affective experiences of those discourses and the underlying language ideologies. The lack of emphasis on banmal in learning materials, while likely motivated by efforts to simplify the language for learners, can actually cause greater confusion and anxiety down the line. Among the 11 interview participants, only three reported actually using banmal, one of whom only used it when talking to herself. This suggests that the way learners are socialized into these discourses, according to their own accounts, may also affect their actual language use.

What we report here as anxiety diverges in some ways from how foreign language anxiety (FLA) has been investigated as a construct in the SLA literature. Horwitz et al. (1986) define FLA as “a distinct complex construct of self-perceptions, beliefs, feelings, and behaviors related to classroom language learning arising from the uniqueness of the language learning process” (p. 128). This broader definition would seem to encompass more ideologically-driven anxiety than the participants in this study report experiencing. Traditionally, FLA has been investigated as anxiety related to speaking or writing in a foreign language. However, in this study, our participants reported that their anxiety came not from the act of speaking Korean itself, but from the idea of speaking banmal with someone and potentially offending them. Some of this anxiety may certainly come from their lack of Korean proficiency in general, and banmal forms more specifically. However, we also want to highlight how discourses around the kind of language expected from L2 speakers can be a major contributing factor to their comfort or discomfort with certain L2 forms. Additionally, learners are often not told that L1 speakers frequently style mix and switch between using various honorific and non-honorific language features, and so they may see any deviation from jondaenmal as a categorical mistake. In that sense, this kind of anxiety may have more to do with a commitment to politeness and a perceived (or actual) lack of pragmatic competence than it does with a construct like FLA.

For Korean language instructors, the results of this study should reaffirm their crucial role in socializing students not only into the linguistic forms of Korean but also into more ideologically-driven aspects of language learning. If banmal is completely absent from their formal learning experience, learners may get the idea that this language form is not for them. As Brown (2016) notes, textbooks’ “portrayal of language learners as cultural novices, with minimal engagement in Korean society, who only ever use polite language, is neither realistic, nor appealing for language learners” (p. 24.) The discourses reported in our study seem to reflect the assumption that instruction in jondaenmal is the responsibility of Korean language instructors, while learners themselves are responsible for figuring out how to use banmal on their own. However, this is at odds with L2 Korean learners’ goals and beliefs about learning. The survey item in our study that received the strongest agreement, “I believe that I need to know how to speak banmal well in order to feel that I am a fluent Korean speaker,” suggests that learning and speaking banmal is important to learners.

Given the findings of this study, there are several pedagogical considerations that may be useful to KFL practitioners. Our first recommendation is to consider integrating both banmal and jondaemal conjugations into grammar instruction at the elementary level. At times, shifting a particular grammar point from jondaenmal to banmal involves only a minor change, and the two styles can be easily introduced together. Rather than viewing these as having to be taught separately, an integrated approach would allow learners to develop their sociolinguistic repertoires from the beginning. There is precedent for such an integrated approach in other L2 settings—for instance, in teaching spoken dialects of Arabic, which is diglossic, to L2 learners (Al-Batal, 2018), including for the purposes of pragmatic development (Nassif & Al Masaeed, 2022). Second, we recommend that KFL practitioners consider incorporating some of the non-traditional methods that learners, especially those who have learned primarily via self-study, are already using to access banmal into their curricula. Brown (2013), for example, showed that Korean dramas are an effective way to teach the pragmatics of non-honorific styles even in the context of a formal classroom. Bringing in authentic posts or videos from social media could be an effective way not only to teach pragmatics but also to engage students by using platforms with which they regularly interact in their daily lives.

Finally, while instructors may, perhaps understandably, expect that learners will pick up some of the conjugations for banmal on their own outside of the classroom, work on interlanguage pragmatics indicates that L2 learners do not necessarily implicitly learn pragmatic norms simply via exposure alone (whether inside or outside the classroom) and often require explicit instruction in this area (see Cohen & Shively, 2007; Michno & Schmitt, 2025). These studies incorporate Schmidt’s (2010) Noticing Hypothesis to account for this phenomenon, which suggests that pragmatic features do not automatically become part of a learner’s interlanguage system; rather, learners need to consciously notice these elements in their input so that input can become intake in the learning process. Michno and Schmitt (2025) point out that explicit instruction can facilitate Noticing, and they call for “awareness-raising activities, broadly speaking, to help students notice and attend to specific pragmatic features and cultural practices, in the hopes of enhancing student understanding and adoption of target norms” (p. 7).

Therefore, we urge KFL practitioners to consider including explicit instruction on the pragmatics of when, how, and with whom to use non-honorific language. Such activities might include the analysis of authentic interactions, role plays, cross-linguistic comparisons and contrasts, and self-reflection activities. Rather than limiting exposure to banmal so that learners do not offend L1 interlocutors, instructors can empower their students to make thoughtful choices about how to use jondaenmal and banmal to fully enact their sociolinguistic agency in Korean.

6. Conclusions and Future Directions

This study examined the self-reported language socialization experiences of L2 Korean learners at a variety of proficiency levels and who studied both in classrooms and through self-study. It highlighted how discourses about language learning may have influenced these experiences, with a specific focus on the non-honorific banmal speech category. To our knowledge, this study is the first to take a language socialization or discourse-ideological approach to this issue. Our study also uniquely contributes to the literature by examining learners’ self-reported affective reactions in connection with their experience of learning the sociopragmatics of Korean. Discourse analysis of interview data, supported by descriptive statistics from questionnaire results, revealed three dominant discourses in the participants’ self-reported language socialization experiences: (1) Jondaenmal is more important for L2 Korean speakers than banmal, (2) banmal does not belong in formal learning settings, and (3) banmal instruction should be delayed to the advanced or intermediate level. The participants also reported affective reactions to these discourses, which had a direct impact on their learning. Out of a strong desire to avoid being impolite, overly familiar, or offensive, some participants reported increased anxiety around making the sometimes complex decision about whether and when to use banmal with their L1 Korean-speaking interlocutors. On the other hand, some participants also reported that when they were able to successfully style-shift to banmal, they experienced greater linguistic confidence and an increased sense of belonging. These affective experiences point to the importance for practitioners of attending to the discourses around honorifics in the KFL space, since such discourses may have a degree of power to influence student learning experiences and outcomes.

While this study sheds new light on participant-reported experiences of KFL discourses around banmal, future studies could take a longitudinal approach to investigate how these affective consequences influence students’ decisions to continue or discontinue foreign language study. Subsequent research could utilize interaction data to see how participants comply with or resist these discourses around banmal in real-time interactions with L1 speakers. Following the work of Li (2010) on L2 Chinese students’ sociolinguistic accommodation to their instructors’ linguistic choices, future work in the area of language socialization could also investigate how L2 speakers are socialized into the use of certain honorific and non-honorific forms in real time across different language use contexts and relationships. Overall, the findings of this study serve as a reminder that non-honorific language forms are crucial for L2 speakers’ ability to enact the sociolinguistic agency necessary to become confident speakers, engage with L1 communities, and negotiate an authentic sense of self in their L2.