Influence of Culture on Categorical Structure of Emotional Words: Comparison Between Japanese and Korean

Abstract

This study targeted individuals from Korea and Japan to investigate whether their concepts of emotion can be classified as more prototypical (central concepts) and less typical exemplars (peripheral concepts), and how they distinguish them. In Study 1, participants listed examples of “emotion,” and the most frequently reported words were defined as central and others as peripheral concepts. Five central and five peripheral concepts chosen from Study 1 were used in the following four studies, which examined the characteristics of the two conceptual classes and how clearly these are distinguished by Japanese and Koreans. Study 2 identified how strongly the central and peripheral concepts were associated with the word “emotion,” and Study 3 evaluated how good these words were as examples of emotion. In Study 4, the central and peripheral words replaced a word “emotion” in a given sentence and participants assessed how natural the sentence sounded. Study 5 evaluated the fuzziness of the boundary between the central and peripheral concepts. The results of all the studies demonstrated that the concepts of emotion for both Japanese and Koreans are organized based on their typicality. The distinction between the central and peripheral concepts is, however, found to be much more ambiguous for Koreans than for Japanese. The present research revealed that there is a common structure for the concepts of emotion, but it also showed that culturally specific categorization styles as well as language variation can influence how people conceptualize emotion.

Keywords

emotion concepts prototype culture Korean Japanese

Introduction

The Prototype Perspective on Emotion Concepts

Concept categorization minimizes cognitive effort by creating rules and guidelines for behavior and decision making (Matsumoto & Juang, 2013; Rosch, 1978). It is theorized that it reduces cognitive load by grouping similar things together, the classic example being “bird” including elements with shared attributes such as “having wings and feathers” and/or “being able to fly.” Real-world complexities created by the likes of penguins and ostriches are simplified by hierarchies and category boundaries setting some elements as more significant than others (Eysenck & Keane, 2000; Rosch, Mervis, Gray, Johnson, & Boyes-Braem, 1976). A prototype perspective suggests these selection criteria develop around perceptually salient “natural prototypes” possessing the most fundamental attributes of the category, making it the perfect example (Rosch, 1973, 1978). Other category members are included if they share a sufficient number of attributes with the prototype. In each category, the prototype can be thought of as located at a central point with distance from that point representing typicality. Rosch (1978) also argues that the hierarchical structure of conceptual categories has three levels, superordinate, basic, and subordinate. Concepts at the superordinate level are related to each other in general terms, for example “vehicles.” The subordinate category with the lowest level of abstraction includes specific items such as “my car.” In between these two is the basic level, which is proposed as the best summation of categories (e.g., car). Basic concepts, the middle level, share the most similarities, and are acquired at a young age and are the most frequently used level in adult life (Eysenck & Keane, 2000; Rosch et al., 1976).

In the field of psychological research, a definition of emotion has been problematic. It has been defined as anything from a physiological reaction to a much more abstract concept, and this range of descriptions is reflected in individuals’ own interpretations of emotional experiences. Fehr and Russell (1984, 1991), however, hypothesized that ignoring an actual definition, the concepts of what people identify as being an emotion or not, is consistent and is based on category rules despite the variations in individual’s emotional experiences, in the same way as they can say what a vehicle is. This has led them to expect that we have a common structure for concepts of emotion, as suggested in the prototype-based model (Rosch et al., 1976).

To investigate whether the prototype perspective applies to conceptual categories related to emotion, Fehr and Russell (1984) set emotion as a superordinate concept and asked their participants to produce words that were examples of emotion. Additionally, their participants rated how typical those words were as emotions. It was found that some of the most frequently reported words, which would be the basic level of categorization, were parallel to what Ekman, Sorenson, and Friesen (1969) have identified as basic emotions, such as happiness, anger, sadness, disgust, and fear. Other emotions such as envy, courage, and pride were also reported, but less frequently. The results suggest that a concept of emotion has a basic-level prototype with all of these words as category members and frequency of reporting representing distance from the prototype. Fehr and Russell (1984) named the frequently mentioned words as central concepts of emotion, which are thought to locate near the prototype at the center of the category, while less frequently reported words were named as peripheral concepts, locating further away. A number of studies conducted by Fehr and Russell (1984) confirmed that the central concepts activate the superordinate word “emotion” more strongly than the peripheral concept words. Also, they are better substitutes to the word “emotion” in given sentences compared with the peripheral concept words.

Although Fehr and Russell (1984, 1991) found that people tend to agree with the central and peripheral classification, they only examined English-speaking participants from Canada, and it is not clear whether the narrowness of this sample means the findings can definitely be extended beyond this group. With this basis, the present research investigated whether the concepts of emotion for Japanese and Korean correspond with the results of Fehr and Russell.

Influence of Culture on Construction of Emotion Concepts

Culture is the localized way of conducting everyday life, and so it must impact heavily on how individuals experience emotion (Averill, 1980; Levy, 1973). One obvious cultural variation is language, and this is a significant factor to define emotion (Russell, 1991). Different languages express emotion differently. This is demonstrated by the terms for emotions in different languages that elude precise translation (Briggs, 1970; Hiatt, 1978; Lebra, 1983; Levy, 1973), and what is classed as emotions also varies across cultures (Matsumoto & Juang, 2013). For example, the Japanese language tends to include words such as “considerate” or “lucky” in the emotion category, but they are generally not considered as emotions in the English language (Matsuyama, Hama, Kawamura, & Mine, 1978). People’s cultural orientation, especially Western individualism with its focus on rights, self-fulfillment, and autonomy and the duty, group, and unity focused Eastern collectivism (Hofstede, 1980; Oyserman, 1993; Schwartz, 1990), can also influence how we perceive and comprehend languages. Indeed, Gudykunst, Yoon, and Nishida (1987) report that collectivistic cultures tend to feel stronger familiarity toward words that indicate their in-group members than those for their out-group members while such tendency is weaker for those from individualistic societies. Such cultural orientations may alter emphasis or frequencies of particular emotion words in people’s daily use, and this can lead some words to be central concepts in one culture but peripheral concepts in another.

Culturally specific categorization styles may also impact the structure of emotion concepts. It is argued that the individualism–collectivism perspectives influence thinking so that those from collectivistic cultures engage in holistic thinking where every element in the world shares a connection, whereas members of individualistic cultures utilize analytic thinking with a world composed of independent objects (e.g., Munro, 1985; Nakamura, 1964/1985; Needham, 1962). Holistic thinking elicits a greater level of attention to context, because a relationship between the item and context is regarded as significant. Holistic thinking also translates to a preference for intuitive and dialectical reasoning, seeking a “middle-way” between conflicting propositions (Fung, 1952; Liu, 1974; Lloyd, 1996; Nakamura, 1964/1985; Nisbett & Miyamoto, 2005; Norenzayan, Smith, Kim, & Nisbett, 2002). In contrast, analytic thinking involves the decoupling of the object from its context, assigning categories based on necessary and sufficient features, and a preference for rules, especially those of formal logic, to explain and predict behavior. This holistic/analytic difference is likely to influence how people distinguish between the central and peripheral concepts.

Mapping this cultural perspective onto Fehr and Russell’s (1984) study, the finding that central concepts are definitively associated with “emotion” more strongly than peripheral ones may be a consequence of their Canadian-only sample. The analytic thinking ability to apply rules may allow them to efficiently classify the two concept types, while participants from a culture with a holistic thinking style that favors the more ambiguous context over rules were not included in the study. For this reason, the present study will target participants from traditionally holistic thinking cultures, Korea and Japan.

Finally, we highlight some similarities and differences between Japanese and Korean cultures. The two are geographically close with similar linguistic systems (Vovin, 1993), and both countries are classified as collectivistic, so should share culturally dependent classification traits. Recently, however, some researchers (Harris, 1997; Mouer & Sugimoto, 1986; Takano & Osaka, 1997) proposed that the emphasis in collectivism varies across East Asia, and even some have claimed that the level of collectivism among Japanese is not very high (Takano & Osaka, 1997). Matsumoto, Kudoh, and Takeuchi (1996) tested Japanese students with an individualism–collectivism scale and found that more than 70% were classified as individualistic. This is supported by Hofstede’s (2001) assessment of his individualism index with Japan’s score of 46 being 22nd highest on individualism out of the 50 countries so far assessed. The most individualistic country is the United States with a score of 90. The same index puts Korea much lower on individualism with a score of 18, placing the country 43rd. If analytic–holistic views influence the concept of emotion, the difference between Westerners and Koreans should be quite noticeable while the differences between Westerners and Japanese may be less pronounced. The distinction between the central and peripheral concepts are expected to be clearer for Japanese than Koreans, with the results for Japanese being more similar to those found by Fehr and Russell (1984).

Present Studies

This article reports five studies conducted with Japanese and Korean university students. They were all native speakers of Japanese or Korean and spend most of their life in these countries. In Study 1, Japanese and Korean participants reported exemplars of the superordinate concept “emotion,” among which frequently reported words were regarded as central concepts and less frequently reported ones were defined as peripheral concepts. The reported words were compared not only against each other but also against the words reported by native English speakers in Fehr and Russell’s (1984) study. In Study 2, participants were presented with the central and peripheral concept words identified in Study 1 and asked to produce superordinate concepts associated with them. Central concepts are expected to be associated with superordinate concept of “emotion” more strongly than peripheral concepts, and so participants would report the word “emotion” more frequently when they see the central words than the peripheral words. Study 3 evaluated how good these central and peripheral words were as examples of emotion with an expectation that central concepts are regarded as better examples of emotion than peripheral concepts. This is further evaluated in Study 4, in which the central and peripheral words replaced a word “emotion” in a given sentence and participants assessed how natural the sentence sounded. Finally, Study 5 examined the fuzziness of the boundary between the central and peripheral concepts. In this study, participants were asked to judge whether the central and peripheral words are emotion or not. If the boundary between the two types of concepts is clear, peripheral concepts are more likely to be excluded from the category of emotion, but a fuzzier boundary would leave them included in the emotion category. All the studies also examined whether the distinction between the central and peripheral concepts differs for Japanese and Koreans. If Koreans engage in holistic thinking more prominently than Japanese, this may manifest as more ambiguous distinction between the two types of concept for Korean individuals than for Japanese.

Study 1: Basic-Level Exemplars of the Concept of Emotion

In this study, we gathered Japanese and Korean basic-level words as exemplars of the superordinate concept “emotion.” According to Fehr and Russell (1984), the words reported with higher frequency are typical examples of emotion and, thus, are central concepts. Less frequently reported words are peripheral concepts, the examples with low level of typicality.

Method

Participants

All participants involved in the present research are Japanese students at Doshisha University, Kyoto, and Korean students at Chung-Ang University, Seoul. In all studies, Japanese students received course credits, while Korean students received 3,000 won (approximately US$3). Written informed consent was obtained from all individual participants. None of the participants took part in more than one study. In all, 201 Japanese (58 male, 143 female, M = 19.05, SD = 1.43) and 207 Koreans (103 male, 104 female, M = 22.6, SD = 2.47) participated in Study 1.

Materials and procedure

The procedure of this study as well as all the following studies in this article strictly followed the procedure used by Fehr and Russell (1984), and so their original questionnaires and instructions in English were translated to be used in the present five studies. First, they were translated to Japanese and Koreans by two native speakers for each group. The Japanese and Korean questionnaires and instructions were then translated back to English by a Korean–English bilingual and a Japanese–English bilingual. After the authors modified the initial Japanese and Korean questionnaires and instructions based on the reverse translations, two Japanese–Korean bilinguals checked the equivalency of these two sets of questionnaires and instructions.

Participants were tested individually in Study 1. They were provided with a sheet of paper with the word “emotion” in their native language (“Kanjou” in Japanese and “Kamzeong” in Korean) written at the top. Participants were asked to list examples of emotion. They were given an example sheet that showed “animal” as the superordinate level and names of species of animals (e.g., rabbits, dogs, tigers) as the basic level. They were instructed to list as many items as come readily to mind during a minute. They were notified when the minute had passed but allowed to continue writing until they were satisfied with their answers.

Results and Discussion

The participants produced a total of 267 different Japanese words and 419 Korean ones. Mean number of words reported by participants were 9.5 (SD = 3.2) by Japanese and 11.3 (SD = 3.4) by Koreans. Number of participants reported each of those words were counted and the 10 most frequently listed words from both groups were identified. The summary is reported in Table 1. To capture the general tendency clearly, we removed words that were reported by only one individual (140 Japanese words and 232 Korean words) from the total number of words. The data reported on Table 1, therefore, are based on remaining 127 Japanese and 187 Korean words. The table also shows words frequently reported by English-speaking participants in the study by Fehr and Russell (1984). The Japanese and Korean words reported on the table correspond to the English translation (in brackets) fairly straightforwardly.

Table 1.

The Top 10 Most Frequently Reported Words by Japanese, Korean, and English Speakers.

Rank	Japanese	%	Korean	%	English^a	%
Rank	Word	%	Word	%	Word	%
1	Kanashii (Sad)	86.6	Kippun (Pleased)	84.5	Happiness	76.0
2	Ureshii (Pleased)	73.6	Sulpun (Sad)	84.1	Anger	74.5
3	Tanoshii (Enjoyable)	72.1	Julgoeun (Enjoyable)	48.3	Sadness	68.0
4	Ikari (Angry)	54.7	Sarang (Love)	46.4	Love	62.0
5	Sabishii (Lonely)	32.8	Uulhan (Depressed)	45.4	Fear	48.0
6	Kowai (Fearful)	28.4	Haengbok (Happy)	42.5	Hate	44.5
7	Tsurai (Painful)	28.4	Hwanan (Angry)	38.2	Joy	41.0
8	Yorokobi (Glad)	26.4	Bunno (Rage)	33.8	Excitement	26.5
9	Kurushii (Agony)	20.4	Ja-jeungnanun (Irritation)	26.6	Anxiety	25.0
10	Iraira (Irritation)	19.4	Joa-haneun (Like)	19.8	Depression	21.0

Note. Japanese, n = 201; Korean, n = 207; Fehr and Russell (1984)^a, n = 200.

Fehr and Russell’s (1984) results by percentage of subjects who listed each of 20 categories of emotion (p. 470).

Table 1 shows that all three groups of participants frequently reported words related to pleasure, sadness, and anger. The Korean list contained two words related to anger, “Hwa” and “Bunno,” which differ in intensity. “Hwa” is the general term for sudden violent displeasure in response to being wronged, whether real or imagined, whereas “Bunno” is vehement, uncontrolled anger. Some differences across countries are observed. For example, English speakers and Koreans frequently mentioned love, but this was not the case for Japanese. Instead, the word “Sabishii” (loneliness) was listed frequently and this may reflect relationship-oriented traits, as Scherer (1997) argued that Japanese tend to conceptualize emotion as a result of interpersonal relationships. Loneliness is commonly experienced when one is missing someone he or she has a relationship with, and thus loneliness is likely to highlight the state of the individual’s relationship more strongly than love, which can be experienced regardless of the state of the relationship (e.g., unrequited love). So, Japanese people may regard such emotion reflecting personal relationships as a highly typical exemplar of the conceptual category of emotion.

Another noticeable difference is that hate was reported by more than 40% of Canadian participants, but it was not frequently reported by Korean or Japanese. Instead, they reported irritation relatively frequently while Canadians rarely did so. Both hate and irritation are a response to encountering something unpleasant, but hate is projected to the object while irritation is held inside. While speculative, this may be because people from Eastern cultures tend to inhibit expressing negative feelings toward the object, whereas people from Western cultures allow themselves to express these feeling toward the source directly (Hofstede, 1980; Oyserman, 1993; Schwartz, 1990).

Overall, Japanese and Korean data are mostly similar with small differences. Japanese participants, for example, reported only three words related to positive emotion while Koreans reported five. The data also did not differ greatly from those for English-speaking participants, indicating that a large number of exemplars for central concepts are shared across different cultures.

Study 2: Eliciting the Superordinate Concept “Emotion” From Basic-Level Exemplars

This study reversed the principal from Study 1 by presenting participants with the basic-level words collected in Study 1 and asking them to list superordinate concepts associated with them. The words representing central concepts would be expected to elicit the superordinate concept word “emotion” (or its equivalent) more easily than the words representing peripheral concepts.

Method

Participants

They were 92 Japanese (30 male, 62 female, M = 19.43, SD = 0.86) and 96 Koreans (19 male, 77 female, M = 23.11, SD = 2.13).

Materials and procedure

Ten words for each language were selected from the basic words generated in Study 1 for use as target items (see the appendix). Half of them were the five most frequently reported items, determined as central concept words. The other five were peripheral concept words, randomly chosen from the remainder of Study 1’s words (excluding the 10 most frequently reported words shown in Table 1). In addition, seven nouns unrelated to emotion were used as filler items.

In accordance with Fehr and Russell (1984), four versions of questionnaires were created. Each contained 10 words. The first and the second versions were composed of three target words set among seven filler items. The third version had four target words and six fillers, and the fourth only had the 10 target words. As the superordinate word for the target words would always be expected to be “emotion,” the filler items were used give the participants some variation in their answers. Number of participants allocated to each version was 23 for Japanese and 24 for Korean questionnaires.

In the study, participants were tested individually, and each was provided with only one version of the questionnaire. They were asked to write one superordinate concept for each of the 10 words on the questionnaire. There was no time limit.

Results and Discussion

In this study, each of the 10 Japanese words received 46 responses (23 from either Version 1, 2, or 3 and 23 from Version 4), and thus total number of responses were 460, among which half (230) was for central concepts and the other half was for peripheral concepts. Similarly, each of the Korean words received total of 48 responses, which resulted in 480 total responses with 240 for each type of concepts. The type of word (central or peripheral) and whether the answer was “emotion” or not were recorded for each response. The percentage of responses being the actual word “emotion” (“Kanjou” in Japanese and “Kamzeong” in Korean) as the superordinate for each term is reported in Table 2. The table also shows the percentage of responses with emotion-related words such as “affect,” “mood,” and “feeling” in the relevant language.

Table 2.

Proportion of “Emotion” and Emotion-Related Words Responses for the 10 Target Emotion Terms.

Japanese						Korean
Central			Peripheral			Central			Peripheral
	Emotion	Related words		Emotion	Related words		Emotion	Related words		Emotion	Related words
Kanashii (Sad)	69.6	4.3	Odoroki (Surprised)	65.2	4.3	Sulpun (Sad)	56.3	22.9	Nollaun (Surprised)	52.1	18.8
Ureshii (Pleased)	69.6	4.3	Tsumaranai (Bored)	50.0	13.0	Kippun (Pleased)	54.2	20.8	Oaeroun (Lonely)	47.9	22.9
Ikari (Angry)	69.6	2.2	Naku (Cry)	26.1	2.2	Sarang (Love)	56.3	12.5	Duryoeun (Fearful)	43.8	27.1
Sabishii (Lonely)	63.0	8.7	Yasashii (Kind)	23.9	2.2	Uulhan (Depressed)	33.3	33.3	Han (Resentment)	33.3	14.6
Tanoshii (Enjoyable)	60.9	8.7	Darui (Languorous)	21.7	15.2	Julgoeun (Enjoyable)	33.3	29.2	Hwalgichan (Vibrant)	14.6	27.1
M	66.52	5.65	M	37.39	7.39	M	46.67	23.75	M	38.33	22.08
SD	41.32	15.59	SD	30.87	12.90	SD	37.21	33.68	SD	33.22	26.49

The results show that all five central concept words in Japanese elicited the word “emotion” much more frequently than the peripheral concepts. This replicates the study by Fehr and Russell (1984) sampling Canadians and demonstrates Japanese have a clear distinction between central and peripheral concepts. In contrast, only three out of the five Korean central concept words elicited the word “emotion” with higher frequency than the peripheral concept words.

The frequency of “emotion” responses for the central and peripheral words were compared using a χ² analysis performed separately for the Japanese and Korean data. The analysis for Japanese data revealed that the type of words was related to the frequency of “emotion” response, χ²(1, n = 460) = 39.10, p < .001, showing that the proportion of “emotion” response between the central and peripheral words (66.52% vs. 37.39%) were significantly different. This demonstrates that Japanese participants associated the word “emotion” with the central concept words more strongly than with the peripheral ones. A similar analysis on the Korean data, however, did not reveal a significant relation between the type of words and frequency of “emotion” response, χ²(1, n = 480) = 3.41, p = .07.

Next, the frequency of “emotion” responses for central concepts were compared between Japanese and Koreans. The result was highly significant, χ²(1, n = 470) = 18.83, p < .001, and that Japanese participants showed stronger tendency to associate central concepts to the word “emotion” than Koreans (66.52% vs. 46.67%). A similar analysis for peripheral concept was not significant, χ²(1, n = 470) = .04, p = .83. These results clearly indicate that the distinction between the central and peripheral concepts is quite vague for Koreans whereas it was robust for Japanese. This is in line with the prediction based on thinking styles; wherein, the more holistic Koreans have a less clear distinction between the two types of concept than the more analytic Japanese.

Similar analysis performed on the frequency of “emotion-related word” responses revealed that the type of words did not influence the dependent variable either for Japanese, χ²(1, n = 460) = .57, p = .45, or for Koreans, χ²(1, n = 480) = .18, p = .67. However, comparison between Japanese and Koreans found the tendency that Korean participants associated all the target words with emotion-related words more frequently than Japanese: central concepts, χ²(1, n = 470) = 30.35, p < .001; peripheral concepts, χ²(1, n = 470) = 20.00, p < .001. In detail, Korean participants produced eight emotion-related words as superordinate concepts for the targets, and they were reported 110 times in total. These words in English translation and their responses are as follows: mood (46 times, 42%), negative emotion (18 times, 16%), positive emotion (13 times, 12%), emotional status (12 times, 11%), affect (10 times, 9%), static emotion (6 times, 5%), feeling (4 times, 4%), and mind (once, 1%). The Japanese only produced three, which were reported 30 times. They were feeling (16 times, 53%), mood (13 times, 43%), and affect (once, 4%).

Overall, the results of this study revealed that Japanese participants consistently associated the 10 targets with the word “emotion” but less frequently with other emotion-related words. This indicates that the Japanese concepts used in this study, both central and peripheral, must have shared some noticeable attributes that enabled them to be easily categorized as emotion, and those attributes are represented well as a concept of emotion. Such representations must have encouraged the participants to consistently regard these 10 basic-level concepts as exclusive members of the emotion category. In contrast, Korean participants did not produce the word “emotion” as frequently as Japanese participants and they associated the target words with emotion-related words more often. This suggests that commonality among the Korean targets was not salient enough for them to be always categorized as emotion. It can also be argued that the core attributes (or properties) of emotions are not represented well for Koreans and they may have only weak agreement about what emotion is. Noticeably, a large proportion (34%) of the emotion-related word produced by Koreans describe valance or dynamism of emotions (positive, negative, and static) while no such tendency was observed for Japanese. This might be because the word “emotion” is too poorly defined to describe some specific situations so people often combine it with other words.

The weaker agreement for the concept of emotion among Koreans was also observable in the produced superordinate word, which were neither “emotion” nor “emotion-related words.” Japanese reported 191 of this kind and Korean reported 166. Of these, 95 (49.74%) indicated objects or actions related to emotional experiences (e.g., child, enjoy, and ghost) and 107 (64.45%) for Japanese and Korean, respectively. The remainder indicated properties of emotion (e.g., adjectives, states, and personality). This suggests that the 10 target words in Korean activated personal experiences more easily than properties of emotion, and thus, the Korean concept of emotion is likely to reflect more individual variations than the Japanese concept.

In summary, the results of this study showed Japanese participants associated the word “emotion” with the central concept words more strongly than with the peripheral ones, this pattern being consistent with English speakers tested in the previous study (Fehr & Russell, 1984). However, this tendency was weaker among Korean participants and their distinction between central and peripheral concepts are much more ambiguous than those from Japan. This is not due to the differences of the target emotion words between the two languages because some central concept words included in both groups (sad, pleased, and enjoyable) were associated with the word “emotion” strongly for Japanese but much less so for Koreans (see Table 2).

Study 3: Prototypicality of the Central and Peripheral Concepts

This study investigated prototypicality of the 10 words in Study 2 as concepts of emotion. The central concept words are expected to be better examples of emotion than the peripheral concept words.

Method

Participants

They were 55 Japanese (28 male, 27 female, M = 19.80, SD = 0.82) and 59 Koreans (six male, 53 female, M = 22.59, SD = 2.26).

Materials and procedure

The 10 words for each language (five central and five peripheral concept words) used in Study 2 were printed on a questionnaire in randomized order. The participants evaluated how well each word would serve as an example of emotion using a scale ranging from 1 (extremely poor) to 6 (extremely good).

Results and Discussion

The mean rating for each word was calculated and reported in Table 3. Similar to Study 2, the central concept words in Japanese were consistently rated as good examples of emotion; however, some central concept words in Korean were rated less highly than some peripheral concepts. A two-way analysis of variance with factors of concept (central vs. peripheral) and culture (Japanese vs. Korean) was performed on the mean “good example” ratings calculated for Japanese and Korean participants. Ratings by male (M = 3.97) and female (M = 4.10) participants were equivalent, and thus, the data from all the participants were combined for this analysis. A main effect of concept was significant, F(1, 112) = 228.87, p < .001, $η_{p}^{2}$ = .67, showing that the central words were judged as better examples of emotion than the peripheral words. A main effect of culture was not significant (F < 1), but a Concept × Culture interaction was present, F(1, 112) = 37.88, p < .001, $η_{p}^{2}$ = .25. A simple main effect of concept was significant for Japanese, F(1, 112) = 218.80, p < .001, $η_{p}^{2}$ = .66, as well as for Koreans, F(1, 112) = 41.73, p < .001, $η_{p}^{2}$ = .27, but the effect was much stronger for the former. A simple main effect of culture was significant for the central concepts, F(1, 112) = 10.54, p < .01, $η_{p}^{2}$ = .09, showing that the ratings for Japanese were higher than that for Koreans. The effect was also present for peripheral concepts, F(1, 112) = 10.64, p = .001, $η_{p}^{2}$ = .09, and this time Korean participants gave higher ratings than Japanese.

Table 3.

Responses Obtained for the 10 Emotion Words Used in Studies 3, 4, and 5.

Japanese								Korean
Central				Peripheral				Central				Peripheral
	Study 3	Study 4	Study 5		Study 3	Study 4	Study 5		Study 3	Study 4	Study 5		Study 3	Study 4	Study 5
Kanashii (Sad)	5.18 (1.03)	4.47 (1.58)	97	Naku (Cry)	2.56 (1.63)	3.43 (1.73)	38	Kippun (Pleased)	5.27 (1.06)	4.00 (1.34)	100	Nollaun (Surprised)	4.03 (1.11)	3.86 (1.39)	90
Ureshii (Pleased)	5.42 (0.85)	4.22 (1.59)	98	Tsumaranai (Bored)	3.51 (1.46)	3.30 (1.74)	81	Sulpun (Sad)	3.73 (1.88)	4.11 (1.43)	100	Oaeroun (Lonely)	3.49 (1.75)	3.80 (1.44)	94
Tanoshii (Enjoyable)	4.78 (1.24)	4.45 (4.73)	98	Yasashii (Kind)	2.75 (1.32)	3.78 (1.69)	54	Julgoeun (Enjoyable)	4.90 (1.04)	3.96 (1.48)	94	Duryoeun (Fearful)	3.58 (1.77)	4.23 (1.37)	96
Ikari (Angry)	4.38 (1.47)	4.59 (2.13)	100	Odoroki (Surprised)	3.49 (1.41)	4.13 (1.63)	92	Sarang (Love)	4.93 (1.41)	3.94 (1.40)	96	Han (Resentment)	3.14 (1.81)	3.63 (1.46)	86
Sabishii (Lonely)	4.71 (1.25)	4.12 (1.59)	97	Darui (Languorous)	3.20 (1.43)	3.45 (1.72)	52	Uulhan (Depressed)	3.25 (1.83)	3.91 (1.45)	96	Hwalgichan (Vibrant)	4.07 (1.55)	3.40 (1.44)	46
M	4.9	4.4	98.1	M	3.1	3.6	62.5	M	4.4	4.0	97.2	M	3.7	3.8	82.4
SD	0.75	1.34	11.20	SD	0.94	0.92	29.73	SD	0.82	0.75	9.04	SD	0.89	0.70	20.91

As expected, the central concept words are regarded as good examples of emotion, and therefore, it is confirmed that they are prototypical members of the emotion category. Overall, the ratings for the peripheral words were significantly lower than those for the central words but the distinction between the two types of concepts were much less clear for Koreans than for Japanese.

Study 4: Substitutability of the Central and Peripheral Concepts Words to Emotion

In this study, each of the 10 words (five central and five peripheral) used in Studies 2 and 3 replaced the word “emotion” in a sentence and participants assessed how natural the sentence sounded. It is expected that replacing the superordinate word in a sentence is easier for prototypical exemplars than for more peripheral ones. In the present study, therefore, the sentence using a central concept word in the place of the term “emotion” should sound relatively natural, whereas the sentence with the peripheral concept word may sound more unnatural.

Method

Participants

They were 400 Japanese (154 male, 246 female, M = 20.07, SD = 3.40) and 400 Koreans (169 male, 231 female, M = 22.54, SD = 2.67).

Materials and procedure

The 20 sentences from Fehr and Russell’s (1984) questionnaire, each of which contained the word “emotion” (e.g., It is hard to be close to someone who shows no emotion), were translated into Japanese and Korean. Two native speakers of each language initially translated the original English text to their native languages and then selected the better of the two translations. The translated questionnaire then had the word “emotion” replaced with one of 20 replacement words. Ten of these replacements came from Study 2, the rest were fillers randomly selected from Study 1. Twenty questionnaires were then produced with each replacement word being inserted into a different sentence for each questionnaire so that no two were the same. Each participant was provided with one of these 20 questionnaires and asked to rate how natural or peculiar each sentence sounds by choosing a number from 1 (very peculiar) to six (very natural).

Results and Discussion

The ratings for each replacement word was averaged across the 20 sentences and reported on Table 3. A two-way analysis of variance with concept (central vs. peripheral) and culture (Japanese vs. Korean) was performed on mean “naturalness” ratings. Ratings by male (M = 3.89) and female (M = 3.97) participants were equivalent, so participants’ sex were not included as a factor in this analysis. The analysis revealed a main effect of concept, F(1, 798) = 110.41, p < .001, $η_{p}^{2}$ = .12, showing that the sentences with the central concept words (M = 4.18, SD = 1.10) were rated as more natural than the sentences with the peripheral words (M = 3.70, SD = .82). A main effect of culture was also significant, F(1, 798) = 4.72, p < .05, $η_{p}^{2}$ = .01, showing that Japanese participants (M = 3.99, SD = 1.2) rated the sentences as more natural than Koreans (M = 3.88, SD = .09). A Concept × Culture interaction was also present, F(1, 798) = 37.15, p < .001, $η_{p}^{2}$ = .04. Similar to Study 3, a simple main effect of concept was significant for both groups of participants—Japanese, F(1, 798) = 137.82, p < .001, $η_{p}^{2}$ = .15; Koreans, F(1, 798) = 9.74, p < .01, $η_{p}^{2}$ = .01—but the effect was larger for Japanese. Japanese participants judged the sentence with the central concept words as more natural than the ones with the peripheral concept words. Although the similar tendency was observed for the ratings by Koreans, the difference between the central and peripheral words was quite small (0.20). A simple main effect of culture was significant for the central concepts, F(1, 798) = 25.30, p < .001, $η_{p}^{2}$ = .03, showing the higher ratings given by Japanese. In contrast, a significant effect for the peripheral concept words revealed that Koreans gave this type of words the higher ratings than Japanese, F(1, 798) = 8.33, p < .01, $η_{p}^{2}$ = .01.

Generally, this study supports the account that more prototypical exemplars are better substitutes for the superordinate word in a given sentence. As in Studies 2 and 3, the distinction between the central and peripheral concepts is consistently more ambiguous for Koreans than Japanese.

Study 5: Fuzziness of the Boundary Between the Central and Peripheral Concepts

This study examined how clear or fuzzy the boundaries between the central and peripheral concepts are for Japanese and Koreans. Participants were presented with 10 words (five central and five peripheral) used in the previous studies and asked whether they are an emotion or not. While central concept words (highly typical exemplars of emotion) are predominantly used to describe emotion, peripheral concept words, due to their lower level of typicality, can be used to describe other things. Therefore, a clear boundary between the central and peripheral concepts would lead to peripheral words being excluded from the conceptual category of emotion. Alternatively, a fuzzier boundary would leave those words included in the emotion category.

Method

Participants

They were 63 Japanese (38 male, 25 female, M = 19.73, SD = 1.17) and 50 Koreans (22 male, 28 female, M = 23.10, SD = 2.32).

Materials and procedure

A total of 50 questions were created, from which 10 were of interest with the remainder being fillers. The research questions asked “Is . . . an emotion?” in which the 10 emotion words used in Study 2 were inserted. Among the 40 filler questions, 20 were composed to elicit definite “no” answers (e.g., “Is the number forty-seven a female?”), 10 to elicit definite “yes” answers (e.g., “Is three an odd number?”), and the rest of the questions would elicit either (e.g., “Is skiing a method of transport?”). Participants answered each question by stating yes or no.

Results and Discussion

Although there were some participants who responded incorrectly to some of the filler questions with obvious answers, we used all participants’ answers for the target questions. The percentage of “yes” response each participant gave for each of the 10 emotion words was calculated and the summary is reported in Table 3. A two-way analysis of variance with factors of concept (central vs. peripheral) and culture (Japanese vs. Korean) was performed on the “yes” responses. Overall, proportion of “yes” responses by male (M = 86.08) and female (M = 84.09) did not statistically differ from each other, so the data from all the participants were combined for this analysis. The analysis revealed a significant main effect of concept, F(1, 111) = 109.88, p < .001, $η_{p}^{2}$ = .50, showing that the central words were regarded as emotion more often than the peripheral words. A main effect of culture was also significant, F(1, 111) = 10.66, p = .001, $η_{p}^{2}$ = .09. Japanese participants were generally more likely to include the words presented in the emotion category than Koreans. A Concept × Culture interaction was present, F(1, 111) = 18.67, p < .001, $η_{p}^{2}$ = .14. A simple main effect of concept was significant for Japanese, F(1, 111) = 123.81, p < .001, $η_{p}^{2}$ = .53, as well as for Koreans, F(1, 111) = 17.03, p < .001, $η_{p}^{2}$ = .13, but the effect was much stronger for Japanese. While Japanese participants strongly favored central concept words as category members over peripheral ones, Korean participants often included both concepts in the emotion category (above 80%), indicating a vague boundary between the two concepts. A simple main effect of culture was only significant for the peripheral concepts, F(1, 111) = 16.04, p < .001, $η_{p}^{2}$ = .13, and Korean participants were much more likely to include those words in the emotion category compared with Japanese.

The results support the account that the boundary between central and peripheral concepts is more ambiguous for Koreans. That many Korean peripheral concept words were included in the emotion category suggests the attributes of the emotion concept is not well represented, and so individual Koreans tend to define emotion more freely than Japanese.

General Discussion

The present study targeted individuals from Korea and Japan to investigate whether the structure of concepts of emotion fits the prototype perspective (Rosch, 1978) and whether the category members can be divided as the central and peripheral exemplars. As in previous studies with Western participants (Fehr & Russell, 1984; Shaver, Schwartz, Kirson, & O’Connor, 1987), the present study with East Asians identified a hierarchical structure of emotion concepts. When the word “emotion” is set at the superordinate level, basic-level words, for example “anger” and “happiness,” are reported as category members. The prototype perspective predicts that basic-level members are organized so that typicality corresponds to location, with highly typical members located at the center of the category and more distant ones being proportionately less typical. The present study supposed that the words reported most frequently by participants are highly prototypical, central concepts, and less frequently reported words are peripheral concepts. The results of Studies 2, 3, 4, and 5 consistently support this disposition.

Study 1 revealed several central concept words that were commonly identified by individuals from Korea, Japan, and Canada (Fehr & Russell, 1984). Candidates for the central concept words are most likely to correspond to basic emotions (happiness, anger, sadness, fear, disgust, and surprise), which are argued to be evolutionally hardwired and culturally invariant (Ekman et al., 1969; Hejmadi, Davidson, & Rozin, 2000; Keltner & Buswell, 1997). The present study found some overlap between central concepts and basic emotions, but there are noticeable exceptions. The basic emotions of “surprise” and “disgust” were not reported as central concepts, while “surprise” and “fearful” were categorized as peripheral by Japanese and Koreans, respectively. On the contrary, nonbasic emotions such as “lonely,” “love,” and “irritation” are frequently reported as good examples of emotion. A reason behind the disparity between basic emotions and central concepts might be that defining basic emotions relies more on physical elements (facial expressions and physiological reactions), while the concept of “concepts” is much more a product of language. As language is a product of culture, linguistic-based definitions are likely to reflect cultural variations (Averill, 1980; Harré, 1986; Lutz, 1980; Russell, 1991; Solomon, 1976). Culture may also be a factor in how often people experience certain emotions, and a frequency of experience could be a result of cultural variation rather than the supposed innate nature of basic emotions.

Although there are cultural variations in the actual words categorized as central and peripheral concepts, we are able to surmise the properties of the two classes. Central concept words identified in the present study have a clear distinction of being positive or negative (see Table 2), whereas a number of peripheral words, for example “surprised,” “vibrant,” and “cry,” can be used for positive and negative situations. This is in line with the argument that being positive or negative is a fundamental property of emotion (Barrett, 2006; Russell & Carroll, 1999), and so can easily distinguish prototypical exemplars. Some peripheral concept words, however, show ambiguous positivity, and instead appear to ascribe the level of emotional engagement more fully than central concepts. Many peripheral words in the present study either indicate engaged emotions, such as “vibrant” and “cry,” or disconnected ones, such as “languorous” and “bored.” These findings suggest that exemplars of emotion categories can describe two major characteristics of emotion, positivity and engagement, and supports the circumflex model of emotion, explaining affective states as arising from two fundamental neurophysiological systems, valence related to a pleasure–displeasure continuum and arousal, or alertness (Posner, Russell, & Peterson, 2005; Russell, 1980).

The present research consistently demonstrated that the distinction between the central and peripheral concepts is somewhat ambiguous for Koreans while it is more clear-cut for Japanese. The boundary between the central and peripheral sphere appears to be more flexible and changeable for Koreans. This difference was in fact expected if the classifications of the central and peripheral concepts reflect cultural-specific cognitive styles of analytic and holistic thinking. We initially predicted that the more individualistic Japanese participants would use an analytic cognitive style to classify the central and peripheral concepts words, and thus would focus on certain rules or properties. At the same time, the more collectivistic Korean people would engage in holistic thinking, leading them to focus on the context instead of the single word. As individual emotion words could be used in a variety of contexts, the distinction between central and peripheral concepts would become ambiguous. Our findings in Studies 2, 3, 4, and 5 supported these predictions.

In addition, the analytic and holistic cognitive styles may alter frequency of experiencing mixed emotion, which could also impact how we distinguish central and peripheral concepts. Barrett, Gross, Christensen, and Benvenuto (2001) report that we distinguish different emotions according to conceptual knowledge of those emotions, and thus individuals who experience each emotion distinctively tend to develop more clearly defined knowledge of each emotion compared with individuals who tend to experience mixed emotions. People with holistic thinking style are reported to experience mixed emotions more frequently than people with analytic thinking (Bagozzi, Wong, & Yi, 1999). In fact, it has been found that when two emotions with contradicting valance (one positive and one negative) are shown, individualistic Americans tend to perceive one of the emotion more predominantly while ignoring the other, but less individualistic Japanese perceived two emotions equally (Peng, Keltner, & Morikawa, 2002). The more ambiguous distinction between the central and peripheral concepts observed for Koreans in the present research could have reflected the tendency that more holistic Koreans experience mixed emotion more often compared with Japanese. Because Koreans experience co-occurrence of positive and negative emotions frequently, their concept of emotion may represent such flexible characteristics rather than defining emotions as something with categorically definable valance characteristics. The present results, therefore, clearly demonstrated that culturally specific cognitive styles influence the concept of emotion. However, we need to be cautious about the interpretation of results as individuals’ cognitive styles were not measured in the present research. Although it is reasonable to assume that Koreans, who are culturally more collectivistic, would engage in a holistic thinking style more often than Japanese, individual variations in thinking styles do not necessarily correspond to individualism–collectivism cultural values measured at country level. Also, the Japanese and Korean words generated in the study did not perfectly match.

The present study defined the central and peripheral words based on how frequently they were reported in Study 1, and obviously the selected words for Japanese and Koreans were different. Although this is an essential procedure in the present research, it produced nonmatching word sets for the two cultures, and thus it is possible that the ambiguous boundary between central and peripheral concepts found among Koreans is due to some particular qualities associated with the Korean words. We assume it is safe to conclude that the property of the central concept is similar for the two cultures because the results of Study 1 showed a number of overlaps between Japanese and Korean word sets. The peripheral words, however, showed much more variations between the two cultures, and it is likely that those peripheral concepts are more culturally specific than the central concepts. For example, the Korean peripheral word “Han” is complex and thought to be unique for Koreans, so that the English equivalent “resentment” falls short of adequately expressing the emotion. Overall, most Japanese and Korean peripheral words in the present research describe engagement of the emotion rather than positivity because many of the peripheral concepts (see Table 1) could be used for positive and negative situations (e.g., surprised, cry, languorous, and vibrant), but there might be other characteristics specific to the peripheral concepts in each culture influencing classification of these two types of concepts. Future research would be necessary to eliminate language variations to further validate the findings of the present study. It may be useful to target bilingual participants from Korea and Japan who are fluent in the same second language and ask them to evaluate a set of emotion words in that language.

In conclusion, the present study revealed some cultural variations in the structure of emotions. The hierarchical structure of emotional concept in both countries can be divided into central and peripheral exemplars. The distinction between the two concepts are, however, more ambiguous for Koreans than for Japanese, and this may reflect the holistic or analytic cognitive styles. It is highly likely that culturally specific cognitive styles alter how people conceptualize emotions as well as how they determine relationships between them. Linguistic variations in emotion words also play a significant role on how people classify central and peripheral concept words. Future research needs to identify examples of these two types of words commonly shared by participants from different cultures. It may also be worthy of consideration to use descriptive situations (e.g., an emotional seen in a film) associated with particular emotion words for the central/peripheral classification instead of presenting the actual words to avoid linguistic specificity.

Footnotes

Appendix

Ten Target Emotion Items Used in Study 2 (the Percentage Indicates the Proportion of Participants Reported Those Words in Study 1).

Rank	Japanese (n = 201)	%	Korean (n = 207)	%
Rank	Word	%	Word	%
1	Kanashii (Sad)	86.6	Kippun (Pleased)	84.5
2	Ureshii (Pleased)	73.6	Sulpun (Sad)	84.1
3	Tanoshii (Enjoyable)	72.1	Julgoeun (Enjoyable)	48.3
4	Ikari (Angry)	54.7	Sarang (Love)	46.4
5	Sabishii (Lonely)	32.8	Uulhan (Depressed)	45.4
6	Naku (Cry)	10.4	Nollaun (Surprised)	15.0
7	Tsumaranai (Bored)	7.5	Oaeroun (Lonely)	13.0
8	Yasashii (Kind)	5.0	Duryoeun (Fearful)	7.7
9	Odoroki (Surprised)	5.0	Han (Resentment)	1.9
10	Darui (Languorous)	4.5	Hwalgichan (Vibrant)	1.9

Note. Central concept words are in bold type.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Eun-Joo Park

References

Averill

J. R.

(1980). A constructivist view of emotion. In Plutchik

Kellerman

(Eds.), Theories of emotion (Vol. 1, pp. 305-340). San Diego, CA: Academic Press. doi:10.1016/B978-0-12-558701-3.50018-1

Bagozzi

R. P.

Wong

(1999). The representation of affect in independent- and interdependent-based cultures. Cognition and Emotion, 13, 641-672.

Barrett

L. F.

(2006). Valence is a basic building block of emotional life. Journal of Research in Personality, 40, 35-55. doi:10.1016/j.jrp.2005.08.006

Barrett

L. F.

Gross

Christensen

T. C.

Benvenuto

(2001). Knowing what you’re feeling and knowing what to do about it: Mapping the relation between emotion differentiation and emotion regulation. Cognition and Emotion, 15, 713-724. doi:10.1080/02699930143000239

Briggs

J. L.

(1970). Never in anger: Portrait of an Eskimo family (Vol. 12). Cambridge, MA: Harvard University Press.

Ekman

Sorenson

E. R.

Friesen

W. V.

(1969). Pan-cultural elements in facial displays of emotion. Science, 164(3875), 86-88. doi:10.1126/science.164.3875.86

Eysenck

M. W.

Keane

M. T.

(2000). Cognitive psychology: A student’s handbook. New Tork, NY: Psychology Press.

Fehr

Russell

J. A.

(1984). Concept of emotion viewed from a prototype perspective. Journal of Experimental Psychology: General, 113, 464-486. doi:10.1037/0096-3445.113.3.464

Fehr

Russell

J. A.

(1991). The concept of love viewed from a prototype perspective. Journal of Personality and Social Psychology, 60, 425-438. doi:10.1037/0022-3514.60.3.425

10.

Fung

Y. L.

(1952). A history of Chinese philosophy ( Bodde

, Trans.; Vol. I). Princeton, NJ: Princeton University Press.

11.

Gudykunst

W. B.

Yoon

Y. C.

Nishida

(1987). The influence of individualism-collectivism on perceptions of communication in ingroup and outgroup relationships. Communications Monographs, 54, 295-306. doi:10.1080/03637758709390234

12.

Harré

(1986). An outline of the social constructionist viewpoint. In Harre

(Ed.), The social construction of emotions (pp. 2-14). Oxford, UK: Basil Blackwell.

13.

Harris

(1997). Culture, people, nature: An introduction to general anthropology (7th ed.). New York, NY: Longman.

14.

Hejmadi

Davidson

R. J.

Rozin

(2000). Exploring Hindu Indian emotion expressions: Evidence for accurate recognition by Americans and Indians. Psychological Science, 11, 183-187. doi:10.1111/1467-9280.00239

15.

Hiatt

L. R.

(1978). Classification of the emotions. In Hiatt

L. R.

(Ed.), Australian aboriginal concepts (pp. 182-187). Princeton, NJ: Humanities Press.

16.

Hofstede

(1980). Culture’s consequences: International in work-related values. Beverly Hills, CA: SAGE.

17.

Hofstede

(2001). Culture’s consequences: Comparing values, behaviors, institutions and organizations across nations (2nd ed.). Thousand Oaks, CA: SAGE.

18.

Keltner

Buswell

B. N.

(1997). Embarrassment: Its distinct form and appeasement functions. Psychological Bulletin, 122, 250-270. doi:10.1037/0033-2909.122.3.250

19.

Lebra

T. S.

(1983). Shame and guilt: A psycho cultural view of the Japanese self. Ethos, 11, 192-210. doi:10.1525/eth.1983.11.3.02a00070

20.

Levy

R. I.

(1973). Tahilians. Chicago, IL: University of Chicago Press.

21.

Liu

S. H.

(1974). The use of analogy and symbolism in traditional Chinese philosophy. Journal of Chinese Philosophy, 1, 313-338. doi:10.1111/j.1540-6253.1974.tb00654.x

22.

Lloyd

G. E. R.

(1996). Science in antiquity: The Greek and Chinese cases and their relevance to the problems of culture and cognition. In Olson

D. R.

Torrance

(Eds.), Modes of thought: Explorations in culture and cognition (pp. 15-33). Cambridge, UK: Cambridge University Press.

23.

Lutz

(1980). Emotion words and emotional development on Ifaluk Atoll. Cambridge, MA: Harvard University.

24.

Matsumoto

Juang

(2013). Culture & psychology (5th ed.). Belmont, CA: Wadsworth Cengage Learning.

25.

Matsumoto

Kudoh

Takeuchi

(1996). Changing patterns of individualism and collectivism in the United States and Japan. Culture & Psychology, 2, 77-107. doi:10.1177/1354067X9621005

26.

Matsuyama

Hama

Kawamura

Mine

(1978). Analysis of emotional words. Japanese Journal of Psychology, 49, 229-232. doi:10.4992/jjpsy.49.229

27.

Mouer

Sugimoto

(1986). Images of Japanese society: A study in the social construction of reality. London, England: Kegan Paul.

28.

Munro

(1985). Individualism and holism: Studies in Confucian and Taoist values. Ann Arbor: Center for Chinese Studies, University of Michigan.

29.

Nakamura

(1985). Ways of thinking of eastern peoples. Honolulu: University of Hawaii Press. (Original work published 1964)

30.

Needham

(1962). Science and Civilization in China: Vol. 4. Physics and physical technology. Cambridge, UK: Cambridge University Press.

31.

Nisbett

R. E.

Miyamoto

(2005). The influence of culture: Holistic versus analytic perception. Trends in Cognitive Sciences, 9, 467-473. doi:10.1016/j.tics.2005.08.004

32.

Norenzayan

Smith

E. E.

Kim

B. J.

Nisbett

R. E.

(2002). Cultural preferences for formal versus intuitive reasoning. Cognitive Science, 26, 653-684. doi:10.1207/s15516709cog2605_4

33.

Oyserman

(1993). The lens of personhood: Viewing the self and others in a multicultural society. Journal of Personality and Social Psychology, 65, 993-1009. doi:10.1037/0022-3514.65.5.993

34.

Peng

Keltner

Morikawa

(2002). Culture and judgment of facial expression. Unpublished manuscript, University of California, Berkeley.

35.

Posner

Russell

J. A.

Peterson

B. S.

(2005). The circumplex model of affect: An integrative approach to affective neuroscience, cognitive development, and psychopathology. Development and Psychopathology, 17, 715-734. doi:10.1017/S0954579405050340

36.

Rosch

E. H.

(1973). Natural categories. Cognitive Psychology, 4, 328-350. doi:10.1016/0010-0285(73)90017-0

37.

Rosch

E. H.

(1978). Principles of categorization. In Rosch

Lloyd

B. B.

(Eds.), Cognition and categorization (pp. 27-48). Hillsdale, NJ: Lawrence Erlbaum.

38.

Rosch

E. H.

Mervis

C. B.

Gray

W. D.

Johnson

D. M.

Boyes-Braem

(1976). Basic objects in natural categories. Cognitive Psychology, 8, 382-439. doi:10.1016/0010-0285(76)90013-X

39.

Russell

J. A.

(1980). A circumplex model of affect. Journal of Personality and Social Psychology, 39, 1161-1178. doi:10.1037/h0077714

40.

Russell

J. A.

(1991). Culture and the categorization of emotions. Psychological Bulletin, 110, 426-450. doi:10.1037/0033-2909.110.3.426

41.

Russell

J. A.

Carroll

J. M.

(1999). On the bipolarity of positive and negative affect. Psychological Bulletin, 125, 3-30. doi:10.1037/0033-2909.125.1.3

42.

Scherer

K. R.

(1997). The role of culture in emotion-antecedent appraisal. Journal of Personality and Social Psychology, 73, 902-922. doi:10.1037/0022-3514.73.5.902

43.

Schwartz

S. H.

(1990). Individualism-collectivism: Critique and proposed refinements. Journal of Cross-Cultural Psychology, 21, 139-157. doi:10.1177/0022022190212001

44.

Shaver

Schwartz

Kirson

O’Connor

(1987). Emotion knowledge: Further exploration of a prototype approach. Journal of Personality and Social Psychology, 52, 1061-1086. doi:10.1037/0022-3514.52.6.1061

45.

Solomon

R. C.

(1976). The passions: Emotions and the meaning of life. Indianapolis, IN: Hackett Publishing.

46.

Takano

Osaka

(1997). “Japanese collectivism” and “American individualism”: Reexamining the dominant view. Japanese Journal of Psychology, 68, 312-327. doi:10.4992/jjpsy.68.312

47.

Vovin

(1993). Notes on some Japanese-Korean phonetic correspondences. In Choi

(Ed.), Japanese/Korean linguistics (Vol. 3, pp. 338-350). Stanford, CA: Stanford Linguistics Association.