John Benjamins Publishing Company his is a contribution from Review of Cognitive Linguistics 11:1 © 2013 John Benjamins Publishing Company his electronic ile may not be altered in any way he author(s) of this article is/are permitted to use this PDF ile to generate printed copies to be used by way of ofprints, for their personal use only Permission is granted by the publishers to post this ile on a closed server which is accessible to members (students and staf) only of the author’s/s’ institute, it is not permitted to post this PDF on the open internet For any other use of this material prior written permission should be obtained from the publishers or through the Copyright Clearance Center (for USA: www.copyright.com) Please contact rights@benjamins.nl or consult our website: www.benjamins.com Tables of Contents, abstracts and guidelines are available at www.benjamins.com A corpus-based approach to emotion metaphors in Korean A case study of anger, happiness, and sadness Ebru Türker Arizona State University1 he major goal of this study is to investigate conceptual emotion metaphors of Korean, particularly those of ANGER, HAPPINESS, and SADNESS, by utilizing a corpus-based analysis he universality of conceptual metaphors continues to be a controversial topic in cognitive linguistics and thus, more cross-linguistic and language-speciic studies are needed to support the theoretical framework of the Conceptual Metaphor heory (CMT) To this end, the current study identiies and examines Korean metaphorical expressions through a conceptual analysis, supported by both quantitative and qualitative methods, and aims to ind out the types of concepts with which ANGER, HAPPINESS, and SADNESS are associated, and thus, to what extent these associations comprise primary (universal) and complex (cultural) metaphors, as suggested by the current view of the CMT I argue that while it is important to distinguish between universal and cultural metaphors, the hierarchical mapping of variation also describes the characteristics of a language vis-à-vis universality or cultural speciicity Furthermore, I claim that the characteristics of metaphorical expressions should also be determined based on analysis of their occurrences in language use he data suggest a positive correlation between frequency and productivity Understanding the frequency and productivity of emotion metaphors through analysis of their occurrence in actual language use will allow better understanding and provide a basis for further investigation of native speakers’ cognitive styles and cognitive tendencies Keywords: anger, conceptual metaphor, corpus linguistics, emotion, English, happiness, Korean, metaphor, metonymy, sadness, submetaphor Review of Cognitive Linguistics 11:1 (2013), 73–144 doi 10.1075/rcl.11.1.03tur issn 1877–9751 / e-issn 1877–976x © John Benjamins Publishing Company 74 Ebru Türker Introduction he major goal of this study is to investigate conceptual emotion metaphors of Korean by utilizing a corpus-based analysis Emotions as well as emotion lexicons for each target domain were determined by a case study conducted prior to the corpus analysis In the case study, three basic emotions, ANGER, HAPPINESS, and SADNESS, and the most representative emotion lexicons were identiied to provide a basis for exploring a large number of metaphorical emotion expressions his study contributes to the literature on conceptual metaphor theory as it relates to the study of emotion metaphors Most of the current research on emotion concepts has focused on English; only a few studies have been conducted on other languages (Barcelona & Soriano, 2004, on Spanish; Kövecses, 2010, on Hungarian; Matsuki, 1995, on Japanese; Mikołajczuk, 1998, on Polish; Yu, 1995, 1998, on Chinese, inter alia) All of these studies apply conceptual metaphor theory within the framework of cognitive linguistics In conceptual metaphor theory, there is a strong belief in the universality of conceptual metaphors, considering that conceptual metaphors are grounded in correlated human thought and experiences, and that they are not language-speciic (Johnson, 1987; Kövecses, 1986; Lakof, 1987; Lakof & Johnson, 1980; Lakof &Turner, 1989) he universality of conceptual metaphors is still an ongoing research topic in cognitive linguistics, and although the inventors of the conceptual metaphor theory (Lakof, 1993; Lakof & Johnson, 1980, inter alia) rarely refer to languages other than English, their claims point to the idea that the most basic metaphors are universal As evidence for such claims, they note that the same conceptual metaphors may exist in diferent languages and function in the same way in the sense that they are grounded in our human physical experience Several important studies have suggested that human emotions, which are abstract in nature, are largely understood and expressed by metaphorical expressions, but the evidence provided in most previous studies comes from the English language (Kövecses, 1986, 1988, 1990, 1991; Lakof, 1987; Lakof & Johnson, 1980; Lakof & Kövecses, 1987) he main idea is that because metaphorical thought has a close relationship with bodily experience and brain function and we think that all human beings share more or less similar experiences, we should expect to see similar metaphors in other languages and cultures around the world However, more cross-linguistic studies are needed to provide further evidence of the universality and variation of conceptual metaphors Furthermore, several scholars (Gibbs, 1994, p 206; Palmer, 1996, p 107) have suggested that universality and cultural variation should be considered not with an “either/or” view, but rather with an integrated and balanced approach hat is, while the bodily bases of language and cognition are more likely to be widespread or universal, the choice of variations from among many possible options within the nature of © 2013 John Benjamins Publishing Company All rights reserved A corpus-based approach to emotion metaphors in Korean bodily experiences depends largely on cultural understanding and interpretation (Yu, 2008, 2009) Gibbs (1999, p 155) indicates that “embodied metaphor arises not from within the body alone, and is then represented in the minds of individuals, but emerges from bodily interactions that are to a large extent deined by the cultural world.” hus, variations in universality and culture-speciicity constitute the general experiential basis of conceptual metaphors based on the interaction between body and culture (Gibbs, 1999, 2003) In this particular study, adhering to the newer version of conceptual metaphor theory that acknowledges the integration of universality and culture, I will be providing supporting evidence from Korean metaphorical expressions of emotions Although this is not a direct contrastive study with English due to the large scope of the analysis, the fact that I will mainly consider Korean emotion metaphors in reference to the emotion metaphors of English as they are proposed in the literature will bring many insights regarding similarities and diferences between the two languages I believe that it is important to highlight such similarities and diferences because these two languages belong not only to diferent language families but also to two distinct cultures, namely East and West he main questions to be answered in this study are (i) what type of metaphorical expressions does Korean have for the concepts of ANGER, HAPPINESS, and SADNESS? and (ii) to what extent the Korean data support the claim that emotion metaphors are universal or cross-linguistically common? In addition to these main questions, the study will also answer two more detailed questions: (iii) to what extent are these emotion metaphors productive and how frequent are they? and (iv) what are the implications of the productivity and frequency of the metaphors in regard to the integration of universality and culture? his paper proceeds as follows: First, I present a primary case study on emotion lexicons of Korean followed by a description of the methodology and the scope of the study Second, I provide the main analysis of emotion metaphors of Korean in the order of ANGER, HAPPINESS, and SADNESS, applying both a corpus-based and an introspective approach In conclusion, I discuss the Korean corpus analysis results in connection with the universal and culture-speciic nature of metaphors Case study In the case study conducted prior to the corpus analysis, I identiied three emotion concepts: ANGER, HAPPINESS, and SADNESS he goals of the case study were threefold: (i) to collect the most representative sample of possible emotionrelated words; (ii) to identify various emotion words, particularly synonyms or © 2013 John Benjamins Publishing Company All rights reserved 75 76 Ebru Türker Table Mean ratings for free-listing emotion terms # of sample # of terms # of happiness terms # of anger terms # of sadness terms # of other terms 26 25.0 5.3 4.8 3.3 9.0 Table Frequency ratings for free-listing emotion terms ANGER hwa, hwanada (anger) silta, sireohada (hate) bunno, buntong, bunnaem, bunhada (rage) jjajeung, jjajeungnada (annoyance) jeungo, jeungohada (infuriated) 20 13 12 12 HAPPINESS gippeum, gippeuda (joyful) jeulgeoum, jeulgeopda (enjoyable) haengbok, haengbokhada (happy) jota (good) jaemi, jaemi-itda (fun) heungbun, heungbundoeda (excitement) gidae, gidaedoeda (expectation) ttatteutada, ttaseuhada (warm) sangkwaehada (refreshing) 24 19 17 10 4 SADNESS seulpeum, seulpeuda (sad) uul, uulham, uulhada (gloom) seoreoum, seoreopda (sorrow) sseulsseulhada (lonely) 26 11 4 OTHERS bukkeureoum (shame) gwichanta (tiresome) dapdaphada (choking) nollada (surprised) museopda (fearful) buranhada (anxiety) sangsilgam (loss) jigyeopda (loathsome) jing-geureopda (disgusting) siljjeungnada (tired of doing) aswpda (feel the lack of) oeropda (lonesome) useupda (ridiculous) jayu (freedom) jwajeolhada (frustrated) honranseureopda (confusion) 10 6 5 5 5 4 4 4 similarly associated ones for one emotion category (e.g happiness, joy for the happiness-related emotion category); and (iii) to determine the frequency of the most commonly used emotion words I replicated Fehr and Russell’s (1984) methodology he participants were native speakers of Korean who had lived in the United States for less than ive years and who had completed a college degree in Korea he participants (N = 26; 11 males, 15 females) were given ive minutes to write down as many emotion-related words, regardless of grammatical category (e.g verb, adjective, noun, etc.), as they could he results of the analysis are provided in Table and Table As Table indicates, the results of the free-listing task show that emotion words related to HAPPINESS, ANGER, and SADNESS are the most prominent and frequent for © 2013 John Benjamins Publishing Company All rights reserved A corpus-based approach to emotion metaphors in Korean Korean native speakers he participants produced an average of 25.0 emotion words each he mean ratings were 5.3 for words in the category of happiness, 4.8 for anger, 3.3 for sadness, and 9.0 for various emotions outside of these three categories Table lists the emotion words in English and the Korean terms2 in italics based on order of frequency he frequency of occurrences of the emotion words assisted me in deciding which lexical items to choose for each target domain for the corpus analysis Overall the words used most frequently by the participants were seulpeum, seulpeuda (‘sad’; 26); gippeum, gippeuda (‘joyful’; 24); hwa, hwanada (‘anger’; 20); jeulgeoum, jeulgeopda (‘enjoyable’; 19); haengbok, haengbokhada (‘happy’; 17); silta, sireohada (‘dislike’; 13); bunno, buntong, bunnaem, bunhada (‘rage’; 12); jjajeung, jjajeungnada (‘annoyance’; 12); and uul, uulham, uulhada (‘gloom’; 11) hese words were additionally tested in the initial stage of the corpora analysis, and the ones that allowed more metaphorical patterns were chosen as the representative lexical items for the target domains Methodology and scope of the analysis For this study, I utilized the Primary Corpus of the Korean National Corpus (KNC),3 which is a collection of modern Korean language in South Korea composed of both written and spoken texts he Primary Corpus includes corpora with various types of annotations; there is a grammatically tagged corpus (15,293, 877 eojul4), a parsed corpus (825,127 eojul), and a semantically tagged corpus (12,691,945 eojul) I have looked at the target lexical items in each of these three corpora, and I decided to use the semantically tagged corpus because it provides the most data of the sort that I need for this study Because conceptual metaphors not always contain speciic lexical items, I chose two lexical items from the free-listing study to represent each target emotion, and then extracted sample occurrences of each item from the Korean corpus In deciding on the target lexical items, I found that more metaphorical expressions occur in the noun category than the verb category; thus I chose nouns for the lexical items For ANGER, I selected two lexical items, hwa and bunno, and their variations (e.g hwanada, hwanaeda, bunto) For HAPPINESS, the two lexical items I chose were haengbok and gippeum, with the variations of gippeu (e.g gippeo, gippeul) And for SADNESS, I selected uul and seulpeum, along with seulpeum’s variations (e.g seulpeo) he analysis was conducted with the following procedures First, I collected all the target lexical items used in the corpus, discarding the homonyms he words or phrases that contain the target lexical items are included, along with a note of their distribution across registers or word classes Here an issue arose as © 2013 John Benjamins Publishing Company All rights reserved 77 78 Ebru Türker to whether or not to code the lexical items as belonging to metaphorical or literal registers As I am dealing with abstract concepts (emotions), the expressions themselves already indicate abstract concepts, and thus the distinction between concrete and abstract usage was not applicable in coding the metaphors herefore, expressions simply verbalizing emotions in a straightforward way, such as hwanasseoyo ‘I was angry’, haengbokhaeyo ‘I am happy’, and seulpeosseoyo ‘I was sad’ were considered to be used literally, and eliminated from the data Such uses manifest basic meanings and not demonstrate any metaphorical mappings Furthermore, the analysis does not consider whether the metaphors under study are explicitly metaphorical (as in simile, analogy, and so on) or implicitly metaphorical (as in substitution and ellipsis) Second, the expressions that were coded as metaphors were then examined in terms of how they were integrated into a context his was done to decide if they represented instances of metaphor as a matter of cross-domain mapping in the conceptual structures that are expressed in language To say that metaphors in language are the product of conceptual cross-domain mapping, their linguistic relation with linguistic expressions, as well as their experiential and cultural motivations, have to be explained in such a way that they can be identiied by the type of inference structures that exist between source and target domains he expressions were checked to see whether the mappings provided a fully-ledged referential baseline between the target and source domains Detailed elucidations of such metaphorical mappings are provided for the metaphorical expressions that are discussed throughout the analysis he semantically tagged corpus provided 2,172 tokens (1,343 for hwa and 829 for bunno) for the concept of ANGER; 4,178 tokens (2,179 for haengbok and 1,999 for gippeum) for the concept of HAPPINESS; and 1,910 tokens (583 for uul and 1,327 for seulpeum) for the concept of SADNESS, including both literal and metaphorical meanings hese tokens yielded 593 metaphorical patterns for ANGER, 952 for HAPPINESS, and 562 for SADNESS Next, I grouped these metaphorical patterns into source domains (e.g UP, FIRE, HEAT, etc.) and for each source domain, the data were further analyzed to ind various features such as entailments or speciications within each domain In the last step, I created a detailed explanation for each source domain, providing hierarchical information, describing the relation with other metaphors and metonymies, and examining the motivation of the mappings by investigating further semantic, cultural, social, and physiological evidence In investigating these three emotional concepts, I witnessed a number of principles of conceptual metaphor theory (Lakof, 1993) One of them is an inheritance hierarchy between mappings he hierarchy allows a number of metaphors to be connected under one general concept and enables complex coherence © 2013 John Benjamins Publishing Company All rights reserved A corpus-based approach to emotion metaphors in Korean across metaphors he reason that I have observed a number of mappings occurring in all three emotional concepts is that these mappings are subsumed under a central metaphor for EMOTION concepts his is in accordance with Kövecses’s earlier works (Kövecses, 1998, 2000, inter alia), where he presented a number of central metaphors of EMOTION and their source domains that apply to all/most/ some emotion concepts Naturally, the hierarchy can also be examined within each emotional concept It is important to note that simultaneous mappings can occur within one emotion concept As Lakof (1993, pp 218–219) points out, it is possible that two diferent parts of a metaphorical expression can make use of two distinct metaphorical mappings at once while focusing on diferent aspects of the target domain, thus creating a simultaneous mapping Likewise, I detected many simultaneous mappings (32 for ANGER, 53 for HAPPINESS, and 14 for SADNESS) in which metaphorical expressions came from more than one metaphorical source domain In short, I observed metaphorical mappings occurring in each of the emotion concepts, namely ANGER, HAPPINESS, and SADNESS, as well as mappings that fell into more than one source domain within a single emotion concept As mentioned earlier, the main purpose of this research is to identify emotion metaphors in Korean, particularly for ANGER, HAPPINESS, and SADNESS Moreover, this study will provide both qualitative and quantitative information about both Korean and English, although the main focus is on Korean I will not only examine similar and diferent metaphors that exist in both languages, but also describe the elaborations and the speciications of the Korean metaphors However, this study will not provide any comparison of the degree of conceptual elaboration or degree of linguistic conventionalization because of the large number of metaphors I am dealing with Neither will I make any quantitative comparison between the two languages, for two reasons (i) Although this corpora analysis allows the identiication of a large number of mappings, it still excludes other emotion metaphors that not contain the target lexical items herefore, future studies should analyze more lexical items in order to make more generalizable claims about the overall structure of the Korean emotion concepts And (ii) the only potentially comparable corpora analysis for English is Stefanowitsch’s (2006) metaphorical pattern analysis he current analysis was conducted in a similar manner; however, the two studies difer in two ways First, Stefanowitsch limited his analysis to a sample of 1,000 occurrences for each target lexical item, whereas I did not set a limit in order to ind the greatest possible number of metaphorical variations Second, Stefanowitsch’s metaphorical pattern analysis considers only one lexical item for each emotion, whereas I chose two lexical items, again to gather as much metaphorical data as I could For these reasons, a quantitative comparison between the two languages is not possible © 2013 John Benjamins Publishing Company All rights reserved 79 80 Ebru Türker Corpora analysis of ANGER for Korean hrough the corpora analysis, I was able to detect 626 metaphorical patterns (658 including simultaneous mappings) for hwa and bunno on the concept of ANGER hese mappings, listed in Table 3, were identiied solely based on these two lexical items, which both can be translated as anger in English I not make any distinction in their English translation because the semantic diference does not matter in this analysis he metaphorical patterns vary in the occurrences of these Table he comparison of ANGER metaphors in Korean and English Anger/being angry is Korean English 10 11 12 13 HOT FLUID IN A CONTAINER FIRE INSANITY AN OPPONENT IN A STRUGGLE A CAPTIVE ANIMAL A BURDEN AGGRESSIVE ANIMAL BEHAVIOR TRESPASSING (CAUSE OF ANGER) A PHYSICAL ANNOYANCE A NATURAL FORCE BEING A FUNCTIONING MACHINE A SOCIAL SUPERIOR INTENSITY ✓ ✓ x ✓ x x x x x ✓ x ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ x ✓ ✓ x ✓ ✓ 14 15 16 17 18 19 20 21 22 23 24 25 A FLUID/SUBSTANCE IN A CONTAINER (UNDER PRESSURE) A LIQUID HEAT/COLD A MIXED OR PURE SUBSTANCE LIGHT DARKNESS HIGH/LOW (INTENSITY) (and DEPTH for Korean) A SLEEPING ORGANISM A DISEASE GORGE A SHARP OBJECT A PLANT ✓ ✓ ✓ x ✓ ✓ ✓ x ✓ x x x ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ 26 27 28 29 PHYSICAL AGITATION LIVING ORGANISM COLOR SOUND ✓ ✓ ✓ ✓ x x x x 30 *EXISTENCE/NONEXISTENCE © 2013 John Benjamins Publishing Company All rights reserved ✓ A corpus-based approach to emotion metaphors in Korean two lexical items Some mappings allow them to be used interchangeably whereas others allow one but not the other he mappings that are listed in (1) to (13) are those that have also been discussed for English (Kövecses, 1986, 1990, 1995, 2000; Lakof, 1987; Lakof & Kövecses, 1987; Stefanowitsch, 2006); those listed in (14) to (25) are additional anger metaphors found by Stefanowitsch (2006) in his corpus analysis he ‘✓’ symbol indicates the existence of the metaphor in the language, while ‘x’ indicates the absence of the metaphor in that language Notice that the mappings listed in (8) and (11) are marked ‘x’ for English, because in Stefanowitsch’s corpus analysis, these mappings were not identiied his does not mean that these metaphors not exist in English; rather, it means that they did not appear when the metaphorical pattern analysis was restricted to one lexical item, in this case, anger here is a possibility of inding such mappings with a diferent lexical item (e.g rage) or even without a lexical item from the target domain he same is true for the Korean metaphors; the listed mappings are based only on the lexical items hwa and bunno he absence of a mapping from this list does not mean that the particular metaphor does not exist in Korean, only that this analysis did not detect it Moreover, I identiied further mappings for Korean, listed in (26) to (29), which are not mentioned in the literature for English he metaphor in (30) is a general metaphor, not a particular metaphor for the concept of emotion he metaphors that I have identiied as occurring in Korean are discussed in Sections 4.1 and 4.2 4.1 Anger metaphors shared by Korean and English he ANGER IS HOT FLUID IN A CONTAINER metaphor [Source: HOT FLUID IN A CONTAINER; Target: ANGER], one of the most commonly found mappings in cross-linguistic studies, also exists in Korean Converging on the metaphoric and metonymic principles, these conventionalized expressions constitute a commonly known systematic conceptualization of ANGER in Korean heir ontological and epistemic correspondences are very similar to those in English as described by Lakof and Kövecses (1987, pp 201–202) In general, the container is considered the body, but there are cases in which the container can be certain parts of the body such as the heart or head (e.g (gaseum mitbadageseo) kkeureooreuneun bunno ‘anger boil up (from the bottom of X’s heart)’, gaseume deulkkeulteon hwa ‘anger seething in X’s heart’) As the temperature of the luid increases, the intensity of the ANGER rises (e.g hwaga kkeulta ‘anger seethe on the simmer’) Much as in English, the luid can also be blood, which corresponds with ANGER (e.g bunnoga (bunnoui pi) deulkkeulta ‘anger (anger’s blood) seethe’, piga meoriro sotkkuchineun bunno ‘blood-raise-to-head anger’) © 2013 John Benjamins Publishing Company All rights reserved 81 ... Benjamins Publishing Company All rights reserved A corpus-based approach to emotion metaphors in Korean Table he comparison of HAPPINESS metaphors in Korean and English Happiness/being happy is Korean. .. expectation he HAPPINESS IS LIGHT mapping [Source: LIGHT; Target: HAPPINESS] is also one of the most prominent metaphors in cross-linguistic studies his metaphor also exists in Korean, in which I have... wholeness in Korean While the HAPPINESS IS A FLUID IN A CONTAINER metaphor is related to a more general metaphor EMOTIONS ARE FLUIDS IN A CONTAINER (Kövecses, 1991, p 34, 2008b, p 135), Korean metaphorical