Gender Representation in the Korean Basic Dictionary for Language Learners through Gendered Personal Nouns

Gender and Learner’s Dictionary

Language and society cannot be separated easily. If society is not yet ready to move away from gender stereotypes, changing the language is the first step toward changing social perceptions (Prechter, 1999, p. 56). A dictionary is a composite expression of the linguistic traditions held by speakers of a given language. A dictionary is also the material reflected by “the evolution of consciousness” (Tenorio, 2000). However, among the various types of dictionaries, the inclusion of cultural elements in learner’s dictionaries is problematic; learner’s dictionaries have low cultural content as their primary goal is to enable the acquisition of the language itself (Rey, 1987).

Active discussions on language and sexism began overseas in the 1970s. While research on the use of sexist language in textbooks was deemed important, a focus on sexist language use in dictionaries was limited (Prechter, 1999). In the 1990s and 2000s, studies examined gendered language and political correctness in learner’s dictionaries (Barnickel, 1999; Busse, 2000; Cowie, 1995; Moon, 2014; Norri, 2019; Prechter, 1999; Tenorio, 2000). The common theme pertained to issues with definitions and example sentences that failed to reflect social change. Linguistic prejudices related to race, gender, and sexual orientation investigated since the 1980s, have gained wider popularity under the theme of political correctness. Language learners, who acquire a community’s language/culture through dictionaries, are considerably affected by political correctness (Barnickel, 1999; Busse, 2000). Therefore, the need to present example sentences that reflect political correctness in learners’ dictionaries is emphasized. Prechter (1999) observed sexist discrimination against women in the definitions and example sentences in major English learner’s dictionaries such as the Collins COBUILD English Language Dictionary (1987), Longman Dictionary of Contemporary English (1978, 1987), and Oxford Advanced Learner’s Dictionary of Current English (1984, 1989). In all three dictionaries, men generally appear as agents in the definitions and example sentences, whereas women are rarely depicted as such. As sexist depictions and a lack of gender awareness have been observed in English learner’s dictionaries, efforts have been made to improve them (Norri, 2019). Dictionary editors and editorial boards require continuous policies to prevent sexist depictions (Tenorio, 2000).

Since the 2000s, with the increase in the number of Korean language learners, various dictionaries for foreigners learning Korean have been developed domestically. Most are in the form of glossaries presenting only headword lists, and the only dictionaries with the structure of a standard dictionary are the Learner’s Dictionary of Korean (2008) and the KBD (2016). The Learner’s Dictionary of Korean (2008), composed of approximately 5,000 basic vocabulary words, was the first dictionary to establish a framework for learner’s dictionaries. However, this dictionary is a small-scale, monolingual Korean-Korean dictionary, available only in print. In contrast, the KBD is a large-scale online bilingual dictionary with approximately 50,000 headwords. However, neither of these dictionaries has specific guidelines for gender-related descriptions in example sentences. Even Won (2011), who presented the framework for KBD example sentences, provided only an abstract guideline. Specifically, content that is not practical, does not consider cultural diversity, and contradicts common sense will not be presented, thereby reflecting the reality of the language. In the most recent study on KBD’s editorial guidelines, Lee (2023) observed only grammatical guidelines for example sentences, with no reference to gender-related guidelines. The KBD was published in 2016 when the issue of gender representation in learner’s dictionaries was already being discussed internationally. Hence, the absence of specific gender guidelines shows that the KBD did not adequately reflect contemporary demands.

Critical Discourse Analysis (CDA)

CDA is a research approach that critically examines the relationship between language and society. It operates on the premise that all discourse is closely linked to social structures and power. Specifically, CDA is a discourse analysis method that investigates how ideology, identity, and inequality are reproduced through texts generated in social and political contexts (Fowler, Hodge, Kress, & Trew, 1979; Van Dijk, 2009; Wodak, 2012).

CDA analysis is conducted in three stages: description, interpretation, and explanation. The description stage focuses on the formal properties of the language within the text, such as vocabulary and grammar as part of the discourse. A text is both a product of the production process and a resource for interpretation. Interpretation refers to both interactional processes among participants and their cognitive processes. This is explained through the concept of “members’ resources” (MR). MR are socially generated and ideologically shaped. The formal properties of a text serve as cues that activate elements of the participants’ MR. Interpretation emerges from the dialectical interaction between these cues and MR. In the explanation stage, discourse is investigated as part of social interaction and as a social practice. Social structures form MR, which, in turn, shape discourse. Thereafter, discourse continuously maintains or alters MR, whereas MR sustain or change social structures (Fairclough, 1996, p. 161). This stage thus explores the relationship between social structures, power, ideology, and discourse.

Interpretation and explanation are closely connected. When MR are employed as interpretative procedures to create and understand texts, they are not only utilized but also reproduced. What connects the interpretation and explanation stages is the reproduction of internalized MR through social structures. This is because the former is concerned with how MR are utilized in discourse processing, whereas the latter focuses on the social construction and transformation of MR, including their reproduction. For this reason, Fairclough (1996) described the combination of interpretation and explanation as aspects of analysis.

CDA focuses largely on the qualitative analysis of texts such as newspaper articles and advertisements. However, there is a need for analysis based on quantitative resources in CDA (Shin, 2018). Discourse analysis using reliable large-scale corpora provides in-depth analyses and determines discourse patterns between language and social phenomena (Baker, 2008). Therefore, this study aimed to analyze the types of gender representation in the example sentences of the KBD, the largest Korean learner’s dictionary, based on the CDA methodology.

The Corpus: Korean Basic Dictionary

The development of the KBD began in 2010, and it was published by the National Institute of Korean Language in 2016. As of July 2024, it contains 51,952 headwords, making it the largest Korean dictionary for language learners. The KBD is the only bilingual Korean learner’s dictionary. This dictionary was created with translation in mind, rendering it the foundational resource for the National Institute of Korean Language’s Korean-Foreign Language Learners’ Dictionary, a multilingual translation dictionary. On the learner’s dictionary website, the KBD and 11 foreign-language translation dictionaries (Arabic, Chinese, English, French, Indonesian, Japanese, Mongolian, Russian, Spanish, Thai, and Vietnamese) are presented side-by-side, allowing users to select a dictionary according to their language. The headword information includes pronunciation, parts of speech, grammatical information, definitions, examples, and idioms/proverbs. The examples do not reflect authentic language use; rather, they are contrived examples constructed by the dictionary editors.

Data Collection and Analysis Procedure

To analyze gender stereotypes and sexist representations in the KBD using CDA, example sentences containing gendered personal nouns were extracted. The Korean vocabulary corresponding to the gendered personal nouns “woman” and “man” is shown in Table 1; each definition follows the Korean Standard Dictionary (2024).

Table 1

Korean gendered personal nouns and definitions

Example sentences containing the four keywords yeoseong, yeoja, namseong, and namja were extracted from the KBD. A total of 5,189 example sentences were extracted, of which 1,907 exhibited gender issues.

Table 2

Number of analyzed examples

The analysis of the examples was conducted in two stages: description, followed by interpretation and explanation. Description is the stage of text analysis focusing on vocabulary and grammar. For vocabulary, sexist words and their repetition were determined. For grammar, the types of processes described in the example sentences, such as actions, events, and attributions, were classified. When expressing situations, participants choose between various grammatical processes and participant types (agent vs. non-agent). These choices are ideologically significant (Fairclough, 1996, p. 120). The criteria for determining actions, events, and attributions followed those of Fairclough (1996). Action refers to a process in which the subject of the sentence (a gendered personal noun) is depicted as the agent (doer) of the action. Event describes “what happened,” specifically referring to cases where the gendered personal noun is the grammatical subject of a passive sentence, containing verbs such as “잡히다” meaning “to be caught” or “강간당하다” meaning “to be raped,” or is the theme of an action, e.g., “간통하다” meaning “to commit adultery” or “바람피다” meaning “to have an affair.” Attribution refers to a process that describes the state, quality, or characteristics of the gendered personal noun.

Based on the analysis of the descriptions, the types of gender representation in the example sentences containing the four terms were further categorized in the interpretation and explanation stages. The results of this categorization enabled the analysis of the MR represented in the dictionary. Additionally, by comparing and observing the ratios of the representation types, the reproduction of MR within the dictionary was described. This provided the foundation for analyzing the relationship between dominant ideologies represented in the KBD, changes in social structures, and related language use.

Description (vocabulary)

Table 3 lists the major co-occurring words and their repetitions observed in the example sentences in which YEOSEONG, YEOJA, NAMSEONG, and NAMJA appear:

Table 3

Co-occurring words with YEOSEONG and YEOJA

Table 4

Co-occurring words with NAMSEONG and NAMJA

Description (grammar)

The sentence types for grammatical analysis presented by Fairclough (1996) were simple sentences. However, because dictionary example sentences include subordination and coordination, the total number of actions, events, and attributions does not match that of the example sentences (see Table 5).

Table 5

Types of process in example sentences containing YEOSEONG, YEOJA, NAMSEONG, and NAMJA

The analysis reveals that the four terms exhibit opposite ratios of actions and events. YEOSEONG and YEOJA are more often the objects of action, whereas NAMSEONG and NAMJA are more often the subjects of action. Additionally, in the case of NAMSEONG, both action and attribution ratios are high, indicating masculinity and superiority in ability.

Interpretation and Explanation

Based on the co-occurring words observed in the description (see Tables 3 and 4), gender representation types can be classified into three categories: (A) Representation of sexist ideology, (B) Representation of departure from sexist ideology, and (C) Representation of post-departure situation from sexist ideology. First, the ratios of YEOSEONG and YEOJA in the three types of representations are as follows (see Table 6):

Table 6

Ratios of the three types of representation for YEOSEONG and YEOJA

In the case of YEOJA, category C rarely appears, and B has a much lower ratio than A, which accounts for 90.5%. On the other hand, for YEOSEONG, while the ratio of A is the highest, there is a relatively even distribution across categories A to C.

Category A <Representation of sexist ideology> is further divided into five subcategories: Victim and perpetrator, Traditional female roles, Interest in appearance, Relationship with men, and Social status and rights issues. The ratio of example sentences for YEOSEONG and YEOJA differs across these subcategories, as shown in Table 7.

Table 7

Subcategories of category A for YEOSEONG and YEOJA

In the subcategory <Victim and perpetrator>, the victim is a woman who suffers (sexual) assault or humiliation or is subject to sexual commodification and derogation, whereas the perpetrator is a woman who frequently engages in violence and insults against others or is portrayed as a femme fatale. YEOSEONG and YEOJA are depicted as victims in 89% and 78% of the cases, respectively, with a much higher ratio of victimization compared to perpetration. The subcategory <Traditional female roles> presents a multifaceted depiction of traditional womanhood. This includes traditional views on marriage, conservative attitudes imposed on women, chastity, subordination to men, preferences against women, and physical inferiority. This subcategory ranks second for YEOSEONG and first for YEOJA, indicating that women’s roles and attitudes are yet largely confined to traditional views. The subcategory <Interest in appearance> highlights societal evaluations of women’s appearance and the importance women themselves place on their looks. Examples of women engaging in cosmetic surgeries, dieting, and following trends are easily observable. In the subcategory <Relationship with men>, two main images of women emerge. First, women are regarded as actively seducing men and placing importance on men’s wealth and abilities in their interactions with them. In contrast, in some instances, men consider women solely as sexual objects. The subcategory <Social status and rights issues> shows a large difference between YEOSEONG at 8% and YEOJA at 1%. This category describes issues related to women’s rights, oppression, and life limitations.

In category B <Representation of departure from sexist ideology>, YEOSEONG accounts for a higher percentage than YEOJA. YEOSEONG includes many expressions representing women’s (physical) autonomy, social participation, and success. On the other hand, YEOJA is depicted in contexts such as “군대를 가다” meaning “to do one’s military service” and “(남자를) 리드하다” meaning “leading (a man),” breaking from traditional female roles.

Category C <Representation of post-departure situation from sexist ideology> reveals situations that women continue to face even after societal advancement. Because there are only two examples in YEOJA, this category is particularly relevant to YEOSEONG. While issues related to the treatment of women are highlighted, it also shows that societal perspectives and systems are evolving to address them.

In the case of NAMSEONG and NAMJA, the majority of examples (92.3%) fall into category A <Representation of sexist ideology>. In contrast, category B <Representation of departure from sexist ideology> accounts for less than 10%, and category C <Representation of post-departure situation from sexist ideology> is nearly absent (see Table 8).

Table 8

Ratios of the three types of representation for NAMSEONG and NAMJA

Category A <Representation of sexist ideology> is broadly divided into four subcategories: Perpetrator and victim, Traditional male roles, Masculinity, and Relationship with women. The ratio of examples for NAMSEONG and NAMJA varies across these subcategories, as shown in Table 9.

Table 9

Subcategories of category A for NAMSEONG and NAMJA

The subcategory <Perpetrator and victim> includes examples where a man harms another person’s life, body, mind, property, or honor. Interestingly, in such cases, the victims are predominantly women. This subcategory also includes instances of sexual harassment, molestation, and sexual assault. In contrast, when a man is depicted as a victim, it is usually in the context of physical harm. The subcategory <Masculinity> includes examples related to society’s perception of “manhood,” featuring examples in which men are expected to be physically stronger than women and more socially capable, which reflects the societal roles imposed on men. In this process, prejudices regarding male superiority are also distinctly evident. The subcategory <Traditional male roles> includes examples that reflect traditional stereotypes about men such as the belief that in a marriage, the man should go out and work to earn money, whereas the woman should manage the household, patriarchal ideas such as the acceptability of having concubines, the preference for the “good wife, wise mother” ideal, and the preference for sons over daughters. Examples also include the characteristics and attitudes expected of men, such as taciturnity and seriousness. The subcategory <Relationship with women> often shows a preference for pretty and young women in relationships with men, as well as the tendency to treat women as sexual objects. In contrast, examples also include cases in which a man is the target of a woman’s promiscuity or infidelity.

In category B <Representation of departure from sexist ideology>, NAMSEONG has a higher percentage than namja. Although the number of examples for NAMJA is nearly seven times greater than that for NAMSEONG, the majority of examples for NAMJA (94%) fall into category A, which includes sexist ideology.

Finally, in category C <Representation of post-departure situation from sexist ideology>, we found only one example showing that men face reverse discrimination after achieving gender equality. The difference in the ratios of category C between YEOSEONG (8.9%) and NAMSEONG (0.9%) is noteworthy (see Discussion).

Based on the CDA methodology, we identified the gender stereotypes represented in the KBD example sentences. In the description stage, by detailing the co-occurring words, particularly verbs and adjectives, we determined not only the presence of gender stereotypes but also attempts to diverge from them. Additionally, examining the ratios of the types of process in each example reveals that women tend to be depicted as non-agents, whereas men are depicted as agents. The description stage is essential for interpretation and explanation analysis, providing a basis for categorizing the types of gender representation. The high proportion of category A shows the prominent use and reproduction of MR related to a social structure dominated by gender stereotypes. However, even though the proportion is relatively lower than that of category A, category B reveals the use and reproduction of MR, which internalizes social changes aimed at overcoming inequality and gender stereotypes. This result clearly demonstrates the characteristics of reproduction which “may be basically conservative, sustaining continuity, or basically transformatory, effecting changes” (Fairclough, 1996, p. 39). A detailed discussion of the results is provided in the Discussion.

Discussion

In this section, based on the analysis of the results, we discuss the maintenance of and changes in gender stereotypes, differences in the representation of women and men, and issues related to dictionary example sentence construction and headword selection. Reflecting on these issues, we propose editorial guidelines that dictionary editors should consider when creating learner’s dictionaries.

First, a high proportion of example sentences includes traditional female and male roles within the category of representation of gender stereotype. Traditional gender stereotypes, patriarchy, and traditional views on marriage are still strongly represented in the example sentences. However, the proportion of NAMSEONG and NAMJA in <Traditional male roles> is higher compared to YEOSEONG and YEOJA in <Traditional female roles>. This finding suggests the tendency to describe men more leniently in patriarchal contexts. Furthermore, women’s roles and attitudes are still confined to traditional views, reflecting a gender-hierarchical society. In particular, in category A, some descriptions belittle women using metaphors: “술과 여자” meaning “alcohol and women,” “남자는 하늘, 여자는 땅” meaning “Men are heaven, women are earth,” and “암탉” meaning “hen” (as in the saying “If a hen crows, the household will fail,” where a hen is used as a metaphor for a woman). Gender stereotypes are also evident in the order of mentioning men and women. The precedence of men in language usage reflects an unequal social perspective that prioritizes them (Hegarty, 2014; Mills, 1995). In examples where YEOSEONG and YEOJA appear, NAMSEONG and NAMJA appear first in 61.5% and 68.5% of the cases, respectively. In the examples in which NAMSEONG and NAMJA appear, the percentages are 60.8% and 70.97%, respectively.

Second, women tend to be depicted as victims and men as perpetrators. In the case of NAMSEONG and NAMJA, the category <Perpetrator and victim> rank second and first, respectively. The category <Victim and perpetrator> ranks first and third for YEOSEONG and YEOJA, respectively. The victim ratios for YEOSEONG and YEOJA are 89% and 78%, respectively, indicating substantially high ratios. Particularly, the proportion of those depicted as victims of sexual objectification is high. For men, their depiction as perpetrators of various forms of physical and psychological violence, including sexual harassment, molestation, and sexual assault, is prominent. In some examples, men are even depicted as violently disrespecting women, such as “아내에게 오줌을 내갈기다” meaning “urinating on his wife.” Moreover, the portrayals of victims and perpetrators can also be analyzed through the types of processes. Table 5 shows that the event ratio is the highest for both YEOSEONG and YEOJA. An event is defined as a process in which the gendered personal noun is the grammatical subject of a passive verb or the object of an action verb. In other words, it not only shows the passive attitude of women but also depicts women as the targets of actions initiated by men as agents.

Third, we found unnecessary and inappropriate references to women in the example sentences. The process of categorizing gender representation types revealed issues in the construction of example sentences and selection of headwords, illustrated by the following examples:

• Ex1) 여자는 순결을 잃어 더러워진 몸이 되고 말았다 (표제어: 더러워지다)
  “The woman became dirty after losing her virginity (headword: become dirty)”
• Ex2) 강간당한 여성 중에는 경찰에 신고를 하지 않는 사람들도 있다면서요? (표제어: 강간당하다)
  “Among the women who are raped, there are some who don’t report it to the police, right? (headword: be raped)
• Ex3) 태반이 미용에 효용이 좋다고 하여 음료와 화장품 등으로 여성들에게 인기를 끌었다 (표제어: 태반)
  “It is said that the placenta is effective for beauty, so it became popular among women as a drink and in cosmetics (headword: placenta).”
• Ex4) 그 노처녀는 이미 마흔이 넘은 과년한 여자였다 (표제어: 과년하다)
  “That spinster was already over forty and too old to get married (headword: be past one’s marriageable age)”
• Ex5) 깨끗하고 청순한 얼굴의 그 소녀가 마을 남성을 후리는 요부라는 것이 믿기지 않는다 (표제어: 요부)
  “It is hard to believe that the girl with the clean and innocent face is a temptress who seduces the men in the village (headword: temptress).”

In Example 1, the meaning of “더러워지다” pertains to losing honor or chastity. Despite having already provided an example sentence for “몸이 더러워지다” meaning “body becoming dirty,” an example with yeoja as the subject is presented. In Example 2, “강간당하다” meaning “be raped” is not even listed as a headword in general Korean dictionaries. The selection of this headword strongly emphasizes the perception of women as victims. Furthermore, the content of the example sentence in the dictionary fails to consider the sensitive experiences of sexual abuse and violence victims. In Example 3, the unnecessary mention of women grooming their appearance is used to explain the headword “태반” meaning “placenta.” In Example 4, the headword “과년하다” meaning “too old to marry” refers to a woman who is past the age considered appropriate for marriage. In modern society, where marriage is no longer an obligation, the term “spinster” is used to demean a woman who marries late. The headword itself fails to reflect the changing times. Other such headwords include “교성” meaning “coquettish voice,” “열녀” meaning “virtuous woman,” and “처녀성” meaning “virginity.” In Example 5, a woman seducing men is described as a young girl. The headword “요부” meaning “temptress” refers to a wicked woman who seduces men. This demonstrates how the image of a young woman is construed in the context of seducing men. These examples demonstrate that unnecessary references to women in example sentences are linked to the headword selection. However, the headword selection principles for KBD presented by Son and Kang (2023) does not include guidelines concerning gender representation. It only mentions that the dictionary should include words widely used by native speakers in everyday life and the vocabulary necessary for Korean language learners. The construction of example sentences and the headword selection in learner’s dictionaries differ from those in general dictionaries (Prechter, 1999, p. 49). It is necessary to select “appropriate” headwords that reflect the current relevance of gender issues and include authentic examples based on corpus data to ensure realism. Additionally, the use of diversified and updated corpus that reflects current societal norms will be indispensable for lexicographers and dictionary compilers.

Fourth, even for the same gender, there are differences in representation ratios. The proportion of category A is higher for YEOJA than for YEOSEONG. In contrast, as demonstrated by the difference in ratios for categories B and C, YEOSEONG shows a clearer perspective on breaking away from gender stereotypes than YEOJA. In category B for YEOSEONG, when combined with nouns, the “YEOSEONG + noun” sequences indicate social success. Additionally, it is modified by adjectives such as “최초” meaning “first,” “세계유일무이의” meaning “the world’s only,” and “성공한 40대” meaning “successful in their 40s.” Although the ratio and distribution are lower compared to those in YEOSEONG, examples related to categories B and C for NAMSEONG do exist. More examples reflecting social change can be found with NAMSEONG than with NAMJA. In category B, only approximately one-third of the examples found for NAMJA compared to NAMSEONG were identified. Overall, the ideology of gender equality and departure from sexist ideology have not been fully realized. Furthermore, in example sentences containing YEOJA and NAMJA, the focus tends to be on personal anecdotes that reflect gender stereotypes. In contrast, YEOSEONG and NAMSEONG appear more frequently in example sentences dealing with social contexts. YEOJA and NAMJA tend to be less formal and more familiar compared to YEOSEONG and NAMSEONG.

Fifth, through category B, a shift in society’s gender-related ideology was noted. Specifically, the use and reproduction of MR reflecting the changing social structure were observed. In the overall ratio, category B accounts for 18.7% for women and 7.6% for men, showing that women are 2.5 times more likely to show a departure from gender-stereotypical ideology than men. Although the proportion of example sentences that reveal gender stereotypes is still the highest for women, positive changes in the societal gender roles expectations, such as social participation, success, and departure from traditional female roles, were revealed. Language shapes ideas (Mills, 1995). If discourse on equality and diversity continues to be emphasized, social perceptions and structures can be changed (Fairclough, 1996). Therefore, because more example sentences in the KBD reflect a shift away from gender stereotypes, it is expected that these examples can also affect the perceptions of learners who absorb Korean culture through the dictionary.

Proposition: Editorial Guidelines for Gender Issues

Generally, learners use dictionaries in nonformal, non-guided learning environments (Prechter, 1999, p. 48). It is essential to provide appropriate information to help foreign language learners, who must decode and encode information such as headwords, definitions, and example sentences, understand social structures. In the case of KBD, which is composed entirely of constructed examples, the role of the dictionary editor is critical. A learner’s dictionary can serve as a mirror reflecting the conscious (or unconscious) social values and judgments of its creators, as well as their interest in major social and ideological issues during the editorial period (Busse, 2000, p. 166; Cowie, 1995, p. 294). However, our findings reveal the KBD’s editorial guidelines lack guidance on language use in the sociocultural dimension. A deliberate effort, such as education and policy-making, is required before the use of gender-fair language can become habitual (Sczesny, Formanowicz, & Moser, 2016, p.8). Therefore, to prevent biased gender representations and reflect the current relevance of gender role depictions, this study proposes the following editorial guidelines, reflecting our findings: