Conceptualization of emotion in Japanese: Adjectives vs. verbs

Full Text

1.   Introduction

While adjectives represent properties and verbs represent actions, emotions can be encoded by both (Croft 1991, Wierzbicka 1995). However, few studies have discussed the semantic differences between such formally corresponding adjective–verb pairs in Japanese. According to Teramura (1982: 140), Japanese emotion adjectives express “emotional states” or “emotional judgments,” whereas emotion verbs convey “temporal movement of mind” or “active movement of emotion.” Nevertheless, this distinction does not address how formally corresponding emotion adjectives and verbs differ in meaning.

It has been pointed out that emotion adjectives are used with a first-person subject, while verbs are used more with a non-first-person subject when they are predicates in the non-past tense (Nitta 1998, Teramura 1982). ¹ However, such differences fall short of accounting for their distinction. Corpus data from the Balanced Corpus of Contemporary Written Japanese (BCCWJ)² suggest that the choice between adjectives and verbs is not strictly determined by the subject. In (2a), the verb kanashimu (grieve) is felicitously used with the first-person subject, whereas in (1b), the adjective kanashii (sad) is felicitously used with the third-person subject.³

(1)  a. Watashi-wa jinrui-no         tameni  kanashimi, mata, haiyū-no   tameni

             1sg-top      humanity-gen for        mourn         also   actor-gen for

             hokoritaku  omou  nodearu.                                          (V)

             proud          think  cop

             ‘I mourn for humanity and also feel proud for actors.’

             (Kishida Kunio, ‘Katarareru kotoba’ no bi)

b. Otona-no  sukikatte-wa        kodomo-wa    kanashii.     (Adj.)

             adult-gen  selfishness-top child-top        sad

            ‘Children are sad about the selfishness of adults.’        (FC2Blog)

These examples highlight the need to examine the semantic differences between formally corresponding Japanese emotion adjectives and verbs.

This paper aims to reveal the differences in the conceptualization of emotion between adjectives and verbs in Japanese. The paper is organized as follows. Section 2 reviews previous studies on the three differences between emotion adjectives and verbs: (i) the cause, (ii) the object, and (iii) the expression. Section 3 presents the methodology used in the analysis, Natural Semantic Metalanguage (NSM). It provides data on the adjective–verb pairs that express emotions in Japanese, along with their frequency of use, as obtained from the corpora. Section 4 offers a semantic analysis and sheds light on the meanings shared by Japanese emotion verbs. Section 5 re-examines the conceptualization of emotion in Japanese from the perspective of cultural attitudes toward emotions. Section 6 concludes the paper.

2.   Previous studies

2.1.  The cause of emotion

Semantic differences between emotion adjectives and verbs have been the subject of discussion in both Russian and English. Contrary to the conventional assumption that the relationship between word class and meaning was arbitrary, Wierzbicka (1995) argued that the differences in word class were systematically correlated with differences in meaning. In (2), although the Russian adjective živ and the verb živu are used in the same sentence, one of which is negated, the sentence is acceptable. The examples in (3a, b), cited from Tolstoy (1985), illustrate that altering the verb to an adjective can render the sentence unacceptable.

(2) Ja živ, no ne živu.

‘I am alive, but I do not live.’

(3)  a. Živu durno.

‘I live badly.’                      (Tolstoy 1985: 58)

b. *Ja živ durno.
‘I am alive badly.’

These examples show the semantic differences between the adjective živ and the verb živu, which are common among adjective–verb pairs in Russian.

Wierzbicka also cited examples from English to support her claim. The following English near-synonyms differ in their meaning, which is related to word class.

(4) a. Mary is worrying about something.
     b. Mary is worried about something. (Wierzbicka 1995: 225)

Examples (4a, b) differ in the conceptualization of the cause of emotion. When the verbal phrase is worrying about expresses a mental action, the cause is interpreted as Mary consciously thinking about something. Conversely, the adjectival phrase (pseudo-participles) is worried about expresses a state in which the cause is interpreted as external. In this case, the feeling is triggered by something.

Another difference is that the words occur with different types of collocations. As verbs refer to a process, verbs of emotion can co-occur with aspectual verbs such as start or stop (Wierzbicka 1995). However, the passive emotional state, expressed by adjectives, lacks emotional boundaries and is less likely to occur with aspectual verbs.⁴ Hence, there are significant differences in the number of observed examples in the corpus. Table 1 presents a summary of the raw frequencies of the verbal phrases {to worry/worrying} about and the adjectival phrases {to be worried/being worried} about that follow the present and past tense forms of the verbs start and stop, based on data from the Corpus of Contemporary American English (COCA).⁵

Table 1 shows that a number of examples of the verbal phrase worry about co-occurring with start or stop were observed, whereas few examples of the adjectival phrase to be worried about were detected. Thus, adjective–verb pairs expressing emotion in Russian and English represent different conceptualizations.

 Table 1. The number of co-occurrences with aspectual verbs

2.2.  The object of emotion

Emotion adjectives and formally corresponding verbs in Japanese differ in the nouns they take as the object of the emotion. They use different case markers when taking the object of the emotion. Japanese emotion verbs are used in the construction “N-o + emotion verbs,” where -o indicates the accusative case. Japanese emotion adjectives are used in the construction “N-ga + emotion adjectives.” While -ga is primarily used as the nominative case marker, it can mark the object as well. In both cases, N represents the object toward which one feels the emotion.

Kato (2001) examined the differences in the object of emotion taken by Japanese emotion adjectives and verbs. Below is a comparison of the acceptability of verbs and adjectives when they take the same noun as their object.

(5) a. Ryokou-o tanoshimu. (V)
        traveling-ACC enjoy
        ‘I enjoy traveling.’
        b. Ryokou-ga tanoshii. (Adj.)
        traveling-NOM fun
        ‘I have fun when traveling.’
(6) a. Yoka-o tanoshimu. (V)
         leisure.time-ACC enjoy
        ‘I enjoy leisure time.’
        b. *Yoka-ga tanoshii. (Adj.)
        leisure.time-NOM fun
        ‘I have fun in my leisure time.’ (Kato 2001: 50)

As in (5a, b), both the verb tanoshimu (enjoy) and the adjective tanoshii (fun) take the noun ryokou (traveling) as the object. However, in (6a, b), yoka (leisure time) can co-occur with the verb tanoshimu (enjoy), yet not with the adjective tanoshii (fun). Kato (2001) attempted to account for the differences in acceptability between these adjectives and verbs based on sentence predictability. As “traveling” is likely to evoke a good feeling, which is enjoyable, the addressee can expect a word that expresses a good feeling to follow the noun. However, whether “leisure time” is enjoyable or not depends on how it is spent. Hence, the use of tanoshii (fun) is unacceptable in (7b) because the addressee cannot sufficiently predict the situation from the sentence.

Thus, Japanese emotion adjectives and verbs differ in the nouns they can take as objects of emotion. Compared to verbs, adjectives are more limited in the nouns they can take as objects.

2.3.  The expression of emotion

Whether the emotions designated by emotion verbs are expressed through action has been discussed at the individual verb level. Although the feelings described by English verbs such as rejoice or grieve are normally expressed through actions such as laughing or crying, it is not impossible to experience these feelings without such actions. This suggests that emotions expressed by English verbs do not necessarily involve physical movement (Wierzbicka 1995).

While there is no general discussion about the expression of emotion with Japanese verbs, some studies have examined the meaning of the verb yorokobu (rejoice). They have made two conflicting claims about its expression of the feeling. Nishio (1993) claimed that the feeling does not always appear on the surface because the feeling cannot be observed by others when it occurs with the adverbial phrase shinchu hisokani (secretly in one’s heart), as follows:

(7) Shinchu hisokani yorokobu.
      in.heart secretly rejoice
         ‘I secretly rejoice in my heart.’ (Nishio 1993: 16)

In (7), the experiencer keeps the feeling inside without revealing it to others. This implies that the feeling expressed by yorokobu (rejoice) is not always expressed in action.

In contrast, Caluianu (2005) stated that yorokobu (rejoice) involved action based on the selective restriction with the adverb hadeni (showily). As this adverb modifies the manner of actions, it can only be used with verbs that involve actions. Consider (8a, b), in which all the verbs express feelings:

(8) a. Hade-ni {yorokobu/ kowagaru/ kanashimu}
    ‘To showily {rejoice/ frighten/ grieve}’
    b. *Hade-ni {konomu/ osoreru/ kirau/ aisuru}
    ‘To showily {like/ fear/ dislike/ love}’ (Caluianu 2005: 8)

While the adverb hadeni (showily) can occur with the verb yorokobu (rejoice), it cannot occur with verbs such as konomu (like) or osoreru (fear). Verbs such as yorokobu (rejoice) or kowagaru (frighten) involve physical actions, such as movements of the hands and feet. However, verbs such as konomu (like) and kirau (dislike) express states of liking or disliking. Hence, based on the acceptability difference, Caluianu (2005) maintained that yorokobu (rejoice) involved action.

Sakaba (2023), summarizing these opposing arguments, concluded that the feeling described by the verb yorokobu (rejoice) was expressed in action, because Nishio’s (1993) counterexamples were accounted for by the principle of coercion. It involved a process of reinterpretation, triggered by “a mismatch between the semantic properties of a selector” (Lauwers & Willems 2011: 1219). Although the feeling designated by yorokobu (rejoice) in (7) cannot be observed by others, it does not contradict the claim that the feeling of yorokobu (rejoice) involves action, as the adverb shinchu hisokani (secretly in one’s heart) temporarily cancels this aspect.

In fact, the adverb shinchu hisokani (secretly in one’s heart) could act as a selector to coerce the inherent semantic properties of a selected element when it occurred with other action verbs. For example, the action described by himei-o ageru (scream) in (9) involved some physical movement to make a loud sound. However, the movement described by the verb was temporarily canceled in the following example:

(9)   Shinchu hisokani himei-o ageta       Juzaburo
          in.heart  secretly  scream-acc raise.pst  Juzabro
          ‘Juzaburo, who screamed secretly in his heart.’   (Akagi Tsuyoshi, Teito tantei monogatari)

This suggests that Juzaburo screamed in his mind, without actually making a sound, to avoid being noticed by others. In contrast, the adverb hadeni (showily) does not cause coercion. As seen in (9b), it does not coerce verbs expressing a state to become verbs expressing an action. This suggests that examining the co-occurrence with the adverb hadeni (showily) can determine whether the emotion expressed by the verb inherently involves action, while the co-occurrence with the adverb shinchu hisokani (secretly in one’s heart) cannot serve as a criterion for this judgment.

Thus, the feeling described by yorokobu (rejoice) is expressed in action unless it occurs with words or phrases that coerce this aspect. Examining the co-occurrence with the adverb hadeni (showily) can determine whether the feeling designated by verbs involves actions.

2.4.  Summary of previous studies

Section 2.1 discussed the differences in the cause of emotion between adjectives and verbs in Russian and English, suggesting that similar patterns may be observed in Japanese. Section 2.2 discussed that the objects of emotion taken by adjectives and verbs in Japanese are different, and attempts were made to account for this difference in relation to predictability. However, why predictability affects the acceptability of adjectives but not verbs has not been examined. Section 2.3 explored the expression of emotion designated by verbs. However, further examination is required to investigate the physical movement involved in emotion verbs in Japanese.

3.   Methodology and data

3.1.  Natural Semantic Metalanguage (NSM)

This study employs the NSM as a framework (Wierzbicka 1996) to capture the meaning of emotion concepts. It describes the meaning of words by decomposing them into a set of the simplest concepts. There are 65 concepts called “semantic primes” (e.g., i, someone, think, feel, etc.) that are considered universal.⁶ The list has been updated based on research conducted in over 30 languages from different language families (Goddard & Wierzbicka 2014, Wierzbicka 2021).

NSM serves two purposes. First, it enables the description of subtle differences in the conceptualization of emotion concepts. Second, it identifies the similarities and differences in meaning across languages through the concepts present in nearly all languages. NSM has been applied in several cross-linguistic studies of emotion concepts (Enfield & Wierzbicka 2002, Gladkova 2010, 2022, Goddard & Ye 2016, Harkins & Wierzbicka 2001, Levisen 2024, Wierzbicka 1999, 2018), including Japanese (Asano-Cavanagh 2014, 2017, Habib & Sakaba 2025, Sakaba 2023).

3.2.  Semantic template

The definitions of words that consist of semantic primes are called “semantic explications.” Some explications are based on the “semantic template,” which is a structure of components shared by explications for words of a particular class or subclass. It allows a precise comparison of words with similar meanings. Consider the following semantic template for emotion adjectives in combination with a copular verb:⁷

(10) Semantic template for English emotion adjectives with the verb “to be”:

          Someone X was happy, angry, sad, … (at this time)

1. this someone thought like this at this time:
2. “____
   ____”                                                                                 [model thought]
3. because of this, this someone felt something (very) good/bad [feeling]
   like people often feel when they think like this                                [typicality]
             (Goddard 2018: 72)

The template consists of three components from (a) to (c), which explain the process by which an emotion is aroused. (a) indicates that the person in question “thought like this,” and (b) corresponds to the content of the thought [model thought]. (c) suggests that someone who thought like (b) felt what people typically feel when they think like that.

While Wierzbicka (1995) provided partial semantic components for English emotion verbs, templates for emotion verbs had not been presented. At that time, the concept of templates had not yet been developed. Hence, this study proposes templates for emotion verbs in Japanese by comparing them with emotion adjectives.

3.3.  Data

Given the vast number of Japanese emotion verbs, this study limits its scope to representative pairs of formally related adjectives and verbs. Japanese adjectives are traditionally classified into two types based on meaning: (i) kanjo keiyoushi (emotion adjectives) and (ii) zokusei keiyoushi (attribute adjectives). However, it is difficult to draw a sharp distinction between them, as is evident from different studies proposing different criteria for categories of emotion adjectives (Kitahara 2010, Murakami 2017, Nitta 1998, Teramura 1982, Yamaoka 2000, 2002).

This study employs Murakami’s (2017: 38) definition that emotion adjectives are “adjectives which can express emotions or senses.” Based on this definition, 88 out of the 642 commonly used adjectives selected from the former Japanese Language Proficiency Test Level 1 were categorized as emotion adjectives. Using NLB (NINJAL-LWP for BCCWJ), the online search system for BCCWJ, verbs morphologically derived from these 88 adjectives were searched. Table 2 presents the nine pairs of emotion adjectives and verbs derived from the adjectives by adding the suffix -mu.⁸

Table 2. Pairs of emotion adjectives and verbs in Japanese

A comparison of the usage frequency of each word shows that the emotion adjectives are more commonly used than verbs. Table 3 lists the raw frequency of pairs of emotion adjectives and verbs, including conjugated and past tense forms, verified by NLB. The relative frequencies in parentheses represent the ratio of each verb and adjective pair, with the total set at 100%, and are rounded to the second decimal place.

The adjectives occur more frequently, except for the verb nikumu (hate).^[9] Relatedly, some verbs that formally correspond to emotion adjectives are no longer commonly used in modern Japanese. In addition to ureshimu (rejoice), hossuru (want) (the verbal form of the adjective hoshii) is found only in proverbs, and the verb suku (like) (the verbal form of the adjective sukida) is no longer in common use (Teramura 1982: 144). Although Japanese has a larger vocabulary of emotion verbs than English, it is still more common to express emotion using adjectives.

Table 3. Frequency of emotion adjectives and verbs in Japanese

4.       Semantic analysis

This section explores the differences in conceptualization between the nine adjective–verb pairs from three perspectives: (i) the cause, (ii) the object, and (iii) the expression. However, examining the differences in the cause of emotion in Japanese is challenging. Unlike English, where co-occurrence with aspectual verbs provides insight, Japanese adjectives do not directly follow the verbs.

This study adopted an alternative approach by examining the differences in the object of emotion between adjectives and verbs to infer the differences in their cause of emotion. As shown in Section 2.2, Japanese emotion adjectives are limited in the nouns they can take as objects compared to emotion verbs.

(11)  a. Yoka-o      tanoshimu. (V)

              leisure.time-acc  enjoy

              ‘I enjoy leisure time.’

b. *Yoka-ga                    tanoshii.  (Adj.)

              leisure time-nom        fun

              ‘I have fun in my leisure time.’        (Kato 2001: 50)

Kato’s (2001) explanation is insufficient to account for the differences in acceptability between (11a) and (11b). Instead, the difference stems from the different conceptualization of the cause of feelings. (11a) is acceptable because the emotion verb tanoshimu (enjoy) presents feelings as action, and its cause lies in the experiencer’s conscious thoughts. Even though the object of emotion does not cause good feelings, the experiencer can have a good feeling through their effort. In contrast, the adjective tanoshii (fun) describes the feelings as a state. As the feeling is dependent on an external cause, (11b) is unacceptable. This accounts for why the verb tanoshimu (enjoy) can be used independently of the nature of the preceding nouns, unlike the adjective tanoshii (fun).

4.1.      The cause and the object of emotion

This section examines differences in the cause and the object of emotion between the adjective–verb pairs in Japanese by searching the corpus for the constructions (“N-ga + emotion adjectives” and “N-o + emotion verbs”). The nine pairs can be roughly divided into three types: whether the feeling is generally considered “good,” “bad,” or “neither” (Table 4).

Table 4. Three types of emotion adjectives and verbs in Japanese

The following subsections focus on illustrative examples from corpus searches where nouns that do not evoke the corresponding feeling, appear as the object of emotion (e.g., nouns that normally evoke negative feelings used as the object of “good feelings”). Within each category, the adjective–verb pairs will be examined in order of frequency.

Japanese linguistic data are mainly taken from the BCCWJ, a balanced corpus of contemporary written Japanese from varied media. However, due to the limited examples, when illustrative examples could not be found, they were taken from the Tsukuba Web Corpus (TWC). It is a corpus of approximately 1.1 billion words, with data from Japanese websites. The examples cited from the TWC are marked as (TWC). Each example includes the author’s name and the book’s title; where the author’s name is unknown, only the name of the website is provided.

4.1.1.  Good feelings

4.1.1.1.    Tanoshii (Adj.) and tanoshimu (V)

The Japanese adjective tanoshii (fun) and verb tanoshimu (enjoy) express positive feelings. The corpus suggests that nouns that evoke such feelings, such as jinsei (life) or tabi (trip), are likely to be used as the object. However, in (13), the noun fujiyū (not free), which normally evokes bad feelings, is used as the object of the verb tanoshimu (enjoy).

(12)  Fujiyū-o    tanoshimu   kurai-no     seishinteki yutori-o

          not.free-acc      enjoy          about-gen   mental      space-acc

          motsukoto-ga, seikōsuru  ryūgaku-no           hiketsu da    to   omoimasu.

          to.have-nom   succeed     study.abroad-gen secret   cop  that think.pol

‘I think having the mental space to enjoy the lack of freedom is the secret to a successful study abroad experience.’ (Chishima Eiichi, Etsugozasso)

(12) serves as a message of encouragement or inspiration to the reader. While “being not free” evokes discomfort, here, it encourages the reader to adopt a positive attitude even in such situations. This positive feeling, caused by the effort of the experiencer, is expressed by the verb tanoshimu (enjoy). Out of 1,237 instances of “N-ga + tanoshii” in NLB, fujiyū (being not free) does not appear at all. Replacing the verb tanoshimu (enjoy) with the adjective tanoshii (fun) in (12) makes the sentence sound unnatural, unless the experiencer is someone who enjoys not being free without effort (Fujiyū-ga tanoshii. “I have fun when I am not free.”).

The adjective tanoshii (fun) is generally used in conjunction with nouns that evoke a sense of enjoyment. In (14), it is employed to describe a feeling triggered by the occurrence of a favorable situation.

(13)  Ironna hito       to     aeru        kara       tanoshii

          various  people  with meet.can because  fun

‘I have fun because I can meet various people’

(Akishima Yuriko, Anaunsa wa naze kietanoka)

The expression kara (because) indicates that the subsequent phrase, “meeting various people,” is the cause of the positive emotion. Thus, the verb tanoshimu (enjoy) can take nouns as the object that normally evoke bad feelings, whereas the adjective tanoshii (fun) is restricted to use with nouns that evoke a sense of enjoyment.

4.1.1.2.    Ureshii (Adj.) and yorokobu (V)

The adjective ureshii (happy) and the verb yorokobu (rejoice) co-occur with nouns that evoke a positive feeling, such as egao (smile) or shōri (victory). In the following example, fukō (unhappiness) appears as the object.

(14)  Tanin-no fukou-o                   yorokonde yūetsukan      ni  hitaru     hito
          others-gen misfortune-acc       rejoice       superiority      in  immerse person
          ‘A person who rejoices in others’ misfortune and immerses in a sense of superiority’ (Oshiete! Goo: TWC)

Though others’ misfortune evokes bad feelings, a person in (15) experiences positive emotions to feel superior. Since this emotion is triggered by a mindset, it is designated by a verb. There are 1,057 instances of “N-o + yorokobu” in NLB, with four instances of fukō (misfortune). In contrast, there are 1,116 instances of “N-ga + ureshii” in the corpus; however, fukō (misfortune) does not occur at all. When fukō (misfortune) is used as the object with the adjective, the sentence sounds unnatural (Fukou-ga ureshii “I am happy about being unfortunate”).

The adjective ureshii (happy) is used with nouns that are generally associated with eliciting positive feelings. In (16), it designates an emotion triggered by an external cause.

(15)   Kyo-wa iikoto           bakaride       ureshii.
          today-top good things     full of        happy
          ‘I am happy that today is full of good things.’   (Yahoo! Blog)

In (16), the occurrence of multiple positive events causes the experiencer to feel a pleasant emotion. Thus, as in the case of tanoshii (fun) and tanoshimu (enjoy), the verb yorokobu (rejoice) can take a noun that evokes a bad feeling as the object, unlike the adjective ureshii (happy).

4.1.2. Bad feelings

4.1.2.1.    Kanashii (Adj.) and kanashimu (V)

The adjective kanashii (sad) and the verb kanashimu (grieve) refer to negative feelings, and nouns that evoke such feelings, such as shi (death) or wakare (separation), are commonly used as the object. However, the noun hito no kanashimi (other people’s sadness) is used in the following example:

(16)    Hito-no kanashimi-o isshoni   kanashimu koto-no dekiru

          person-gen sadness-acc together grieve         to-gen   can

          ningen-ni    narou     to     omoimashita.

          person-dat become  that   think.pol.pst

          ‘I thought I would become a person who can grieve together with others.’

          (Setouchi Jakucho, Jakuan dayori)

The sorrow of others who are not related to oneself does not affect people. (17) expresses the author’s desire to become a kind-hearted person who can feel sorrow for others. This feeling is caused by one’s conscious thought and is expressed by the verb kanashimu (grieve). Though there are 215 instances of “N₂-ga + kanashii” in NLB, there are no instances of a noun that does not evoke such feelings used as the object. Replacing the verb kanashimu (grieve) with the adjective kanashii (sad) results in an unnatural expression (Hito no kanashimi-ga kanashii “I feel sad about other people’s sadness”).

The adjective kanashii (sad) is used with nouns that evoke sadness. In the following example, it describes an emotion triggered by an external cause.

(17)  Yatto aeta  noni,     anna        fuu-ni     wakarete  shimau nante… tottemo kanashii.

         finally  met  though  that.kind  way-dat  separate  end.up  such      very      sad

       ‘Even though we finally met, to part in such a way… I’m so sad.’

       (Kashiwa Tetsu, Aishiteru monogatari)

The separation from someone the experiencer met after a long time brings a strong emotional reaction. The adjective kanashii (sad) describes this emotion, which forcefully emerges due to an external cause. Thus, the verb kanashimu (grieve) can take nouns as the object that do not evoke sadness, while the adjective kanashii (sad) does not.

4.1.2.2.    Kurushii (Adj.) and kurushimu (V)

The adjective kurushii (distressed) and the verb kurushimu (suffer) take nouns that evoke these feelings as the object. However, the following example takes a noun that does not cause this feeling, kurushimanakuteii koto (things they do not need to suffer for):

(18)   Kurushimanakuteii koto-o        wazawaza     kurushindeiru  nodesu.

           suffer.neg.don’t have to  thing-acc   deliberately  suffer.prog        cop.pol

           Oya-no       omoikomi-ga, muyouna   kurushimi-o   tsukuridashiteiru nodesu.

           parent-gen belief-nom    unnecessary suffering-acc creat.prog      cop.pol

           ‘They deliberately suffer for things they do not need to suffer for. The beliefs of the parents create unnecessary suffering.’ (Oya ryoku shūchū kōgi: TWC)

The author believes that parents suffer from emotions they do not need to suffer because of their misconceptions. The verb kurushimu (suffer) describes the feeling caused by their conscious thoughts. There are 464 instances of “N-ga + kurushii” in NLB; however, none use a noun as in the above example. Replacing the verb with the adjective in (19) results in an unnatural expression (Kurushimanakuteii koto-ga kurushii “I feel distressed by things I shouldn’t have to suffer through”).

The adjective kurushii (distressed) is used with nouns that are commonly associated with suffering, as follows:

(19)    Netsu-ga tsuzuite,  darukute    kurushii.

          fever-nom  continue  sluggish     distressed

          ‘The fever continues, and I feel sluggish and distressed.’

          (Itō Eiki, Hito wa shineba gomi ni naru)

The experiencer is placed in a physically and emotionally taxing situation due to a prolonged fever, which imposes a sense of suffering. In this context, kurushii (distressed) is used for a feeling caused by fever. While the verb kurushimu (suffer) can take objects that do not inherently evoke emotional pain, the adjective kurushii (distressed) is used with nouns that are semantically associated with suffering.

4.1.2.3.    Kuyashii (Adj.) and kuyamu (V)

The adjective kuyashii (regretful) and the verb kuyamu (regret) take nouns that evoke such emotions as the object. Nevertheless, the noun arayurukoto (everything) is used in the following example:

(20)   Fumanda to   omou hito-wa,      jibun-no       jinsei-o  kōkaisuru   bakaride,

          dissatisfied that think person-top  oneself-gen   life-acc lament  only 

          arayurukoto-o   kuyande bakari iru.

          everything-acc regret      only    be

          ‘Those who think they are dissatisfied only lament their own lives and always regret everything.’ (Agawa Sawako, Agawa sawako no omiaihōrōki)

People do not regret everything because it would hinder their ability to live their lives. The author argues that people dissatisfied with life regret everything because of a mindset that focuses only on the negative. They would not have this feeling if they had the right mindset. As the emotion is believed to be caused by the experiencer, the feeling is expressed by a verb. In NLB, there are 137 instances of “N-ga + kuyashii,” in which there are no objects like (21). Replacing the verb with an adjective results in an unnatural expression (Arayurukoto-ga kuyashii “I am regretful about everything”).

The adjective kuyashii (regretful) is used with nouns that denote situations likely to evoke regret or frustration:

(21)  Sensei-ga nani-o        itteirunoka   rikai                 

         teacher-nom    what-acc   saying           understanding  cannot.do

         Wakaranai        koto-ga       kuyashii.

         not.understand  thing-nom  frustrating

         ‘I cannot understand what the teacher is saying. I feel frustrated that I do not understand.’ (Matsuoka Yayoi, Itte kimasu!)

The experiencer has negative feelings because she cannot understand the teacher’s remarks. This places the experiencer in a state where they feel the emotion designated by the adjective kuyashii (regretful). While the verb kuyamu (regret) can take objects that do not inherently evoke such feelings, the adjective kuyashii (regretful) co-occurs only with nouns that are semantically associated with regret or frustration.

4.1.2.4.    Nikui (Adj.) and nikumu (V)

Both the adjective nikui (hateful) and the verb nikumu (hate) are used in the following example, which quotes a phrase used as a teaching about how one should not hate the sinner.

(22)   “Tsumi-o nikunde hito-o nikumazu” to   iu  ga, sonoyōna koto-ga kanō nanoka.

           sin-acc hate       person-acc hate.not     that say but  such   hing-nom possible cop

          “They say ‘hate the sin, not the sinner,’ but is that possible?”

           (Horiuchi Issei, Kokoro no megusuri)

The phrase intends to consciously direct emotion toward the sin to suppress hatred toward the criminal. This emotional state is subject to the experiencer’s conscious thought and is lexicalized by the verb nikumu (hate). Replacing it with the adjective nikui (hateful) reduces its acceptability (Tsumi-ga nikui “Sin is hateful”).

The adjective nikui (hateful) is used with nouns that denote situations or entities likely to provoke hatred. In the following example, the experiencer is overtaken by a sense of hatred, which is described by the adjective.

(23)   Hito-no ii             fuufu-o              kurushimeteiru   yatsura-ga  yappari  

            person-gen  good       couple-acc      tormenting         guys-nom    after.all   hateful

          ‘I find those who are tormenting the kind-hearted couple hateful.’

          (Izumiya Shigeru, Togi sumasareta rokkan)

The experiencer finds themselves overwhelmed by negative feelings toward people who cause suffering to good people. While the verb nikumu (hate) can take objects that do not inherently evoke such emotions, the adjective nikui (hateful) co-occurs only with nouns that are semantically associated with hatred.

4.1.3.  Neutral feelings

4.1.3.1.    Natsukashii (Adj.) and natsukashimu (V)

The adjective natsukashii (nostalgic) and the verb natsukashimu (miss) take nouns that describe a place or time familiar to the experiencer, such as “hometown” or “era.” However, in the following example, otozureta koto no nai basho (a place I have never been to) is used as the object.

(24)  Kono ongaku-o kiki    nagara, otozureta  koto-no     nai     basho-o,

this    music-acc listen while     visit.pst   thing-gen  not    place-accboku-wa nandomo       natsukashimu.

1sg-top many.times    miss

          ‘While listening to this music, I always miss the place I have never visited.’

          (From seaside: TWC)

The feeling is not evoked by the place because the experiencer has never visited it. Through music, the experiencer consciously generates the feelings of the place. This feeling is expressed by the verb natsukashimu (miss). In NLB, there are 208 instances of “N-ga + natsukashii.” However, there are no examples like the one mentioned above, and replacing the verb with an adjective makes the expression unnatural (Otozureta koto no nai basho-ga natsukashii “I feel nostalgic for a place I’ve never been.”). The adjective natsukashii (nostalgic) is used with nouns that evoke nostalgia.

(25)  Sarasaratoshita suna-no   tezawari-wa,  kodomo-no koro-no

           smooth                sand-gen  texture-top   child-gen    time-gen

           kioku-to          musubitsuiteite,         naniyara     natsukashii.

           memory-with  connected                          somehow    nostalgic

          ‘I feel nostalgic when I touch the smooth sand, as it is connected to childhood memories.’ (Nihon sazō renmei, Sandokurafuto nyūmon)

The tactile sensation of touching sand evokes childhood memories and elicits a sense of nostalgia, expressed through the adjective natsukashii (nostalgic). While the verb natsukashimu (miss) can take objects that do not inherently evoke nostalgic feelings, the adjective natsukashii (nostalgic) co-occurs only with nouns semantically associated with nostalgia.

4.1.3.2.    Urayamashii (Adj.) and urayamu (V)

The adjective urayamashii (envious) and the verb urayamu (envy) often take a person who possesses qualities admired by the subject as the object. In the following example, the phrase shinda monotachi (those who died) is used as the object.

(26)  Kannan-wa tsuzuki,   soshite karera-wa kurikaeshi,

          harshness-top continue and      3pl-top     repeatedly

          shinda   monotachi-o urayanda nodatta.

          die-pst  people-acc   envy.pst  cop.pst

          ‘The harshness continued, and they repeatedly envied those who had died.’

          (Aoyama Masahide, Saigo no kiseki)

Generally, the deceased are not objects of envy. However, kannan (harshness) indicates that the reality is so harsh that the experiencer envies the deceased, who cannot feel this harshness. This feeling is described by a verb because the cause of this emotion is the mental state of the experiencer who envies the dead. Replacing it with an adjective reduces its acceptability (Shinda monotachi-ga urayamashii “I am envious of those who died”).

The adjective urayamashii (envious) is used with nouns that denote situations or attributes likely to evoke envy. Consider the following example:

(27)   Hada-ga kireina       no-ga      urayamashii.

           skin-nom  beautiful   that-acc  envious

           ‘I feel envious of her beautiful skin.’  (Shibata Yoshiki, R-0 amour)

The experiencer directs a gaze of envy toward another person’s beautiful skin and finds themselves unable to suppress the feeling of envy. While the verb urayamu (envy) can take objects that do not evoke such emotions, the adjective urayamashii (envious) co-occurs only with nouns that are semantically associated with envy.

4.1.3.3.    Oshii (Adj.) and oshimu (V)

The adjective-verb pair oshii (disappointed) and oshimu (yearn) takes objects such as life or parting. In (29), the noun shinshoku (sleep and meals) is used as the object of the verb oshimu (yearn).

(28)   Shinshoku-o oshinde shuzaisuru hodo-no      ōkina jiken

          sleep.and.meals-acc   yearn    report         extent-gen  big     incident

‘Major incidents that necessitated sacrificing sleep or meals’

 (Hokkaidōshinbun shuzai han, Nihon keisatsu to uragane)

Since sleep and eating are essential activities, they are not regarded as actions to regret. In (29), the author, a journalist, considers incidents so important and urgent that he feels his time is too valuable to be spent on essential activities. Since this feeling is caused by conscious thought, it is designated by the verb oshimu (yearn). Replacing it with an adjective reduces its acceptability (Shinshoku-ga oshii “I find it regrettable to spend time on sleep and meals”).

The adjective oshii (disappointed) is used with nouns that represent valuable or meaningful entities whose loss is likely to evoke emotional reluctance or regret. Consider the following example:

(29) Atarashii kutsu  demo  nai  kedo,     suteteshimau           no-wa     oshii.

           new           shoes  even   not  though  throw.away           that-top  regretful

          ‘I feel regretful about throwing them away, even though they’re not new shoes.’ (Tanaka Komimasa, Yashi no tabi)

The experiencer is hesitant to discard a pair of shoes because they are still usable. The usefulness of the shoes gives rise to a sense of hesitation described by the adjective oshii (disappointed). While the verb oshimu (yearn) can take objects that do not evoke such emotions, the adjective oshii (disappointed) co-occurs only with nouns that are semantically associated with such feelings.

4.1.4.  Summary

This section examined adjective–verb pairs that expressed “good,” “bad,” and “neutral” feelings, focusing on the types of nouns they take as objects. Verbs expressing “good feelings” take nouns that evoke negative emotions, reflecting the belief that positive attitudes toward negative experiences enhance life satisfaction. In contrast, there is little incentive in embracing negative feelings; therefore, verbs expressing them rarely take positive stimuli as objects. Overall, emotion verbs are more flexible in co-occurring with nouns that evoke opposite or unrelated feelings than emotion adjectives.

The difference in the objects that verbs and adjectives can take reflects the difference in conceptualization of the cause of emotion. Emotion verbs can take a wider variety of nouns as objects because the emotion is caused by the experiencer, and the type of emotion is independent of the object. In contrast, adjectives cannot co-occur with such nouns because the emotion is triggered by an external cause, and the type of emotion is dependent on the object. Thus, the difference in conceptualization of the cause of emotion between adjectives and verbs is also observed in Japanese.¹⁰

4.2.      The expression of emotion in Japanese emotion verbs

Section 2 showed that the feelings described by the Japanese verb yorokobu (rejoice) involved physical movement. As the verb kanashimu (grieve) can occur with the adverb hadeni (showily) per Calaianu (2005), it is considered to be expressed in action.

This section examines whether the remaining seven verbs (excluding yorokobu (rejoice) and kanashimu (grieve)) involve movement based on the co-occurrence with the adverb hadeni (showily). As shown below, some verbs can co-occur with it, while others cannot.

(30) a. Hade-ni {tanoshimu/ kurushimu/ kuyamu}

‘To showily {enjoy/suffer/regret}’

b. ??Hade-ni {nikumu/ natsukashimu/ urayamu/ oshimu}

‘To showily {hate/miss/envy/yearn}’

The feelings designated by tanoshimu (enjoy), kurushimu (suffer), and kuyamu (regret) involve physical movement. In contrast, the feelings described by the remaining verbs do not always manifest on the surface. They can be contained within an internal state, allowing the individual to experience them without being noticed by others. Hence, Japanese emotion verbs differ in how the emotions are manifested. Among the adjective–verb pairs examined in this study, half of the verbs involve physical movement, while the others do not.

4.3.      Semantic template for Japanese emotion verbs

This study explored the differences in the conceptualization of emotion between adjective–verb pairs, focusing on the cause, object, and expression of emotion. Section 4.1 argued that emotions expressed by adjectives were the result of an external cause, while those expressed by verbs stemmed from the experiencer’s conscious thought. Section 4.2 showed that emotions described by some Japanese verbs involved physical movement, while others did not. Based on these findings, the study proposes the following template for Japanese emotion verbs:

(31) Semantic template for Japanese emotion verbs;

Someone X tanoshinda, yorokonda… (at this time)

a. this someone wanted to do something

because of this, this someone thought like this at this time:

b. “____
____” [model thought]

c. because of this, this someone felt something (very) good/bad [feeling]

like people often feel when they think like this [typicality]

d. this someone wanted to do something because of this [expression]

This shares the basic structure of the template for English emotion adjectives, in which the experiencer feels a certain emotion as a result of a model thought. However, they differ in the cause and the expression of emotion. Component (a) states “this someone wanted to do something” and “because of this, this someone thought like this at this time,” representing the cause of emotion where the experiencer holds the thought by their will. The feeling is the result of the thoughts voluntarily directed at a certain object.

Furthermore, some Japanese emotion verbs involve physical movement, while others do not. As discussed in Section 2.3, the feeling expressed by English verbs, such as rejoice, often involves physical movement, although not always. The NSM explications for these words include the component “X wanted to do something because of that,” which invites the inference that the feeling is expressed in action (Wierzbicka 1995: 226). This semantic component is employed as component (d), as the feeling designated by Japanese verbs is not always expressed in action.

To illustrate the applicability of this template, we propose the following explications for the adjective kanashii (sad) and the verb kanashimu (grieve):

(32) Someone X was kanashii (at this time)

a. this someone thought like this at this time:

b. “something bad happened
                     I don’t want this”

c. because of this, this someone felt something bad
                     like people often feel when they think like this

(33) Someone X kanashinda (at this time)

a. this someone wanted to do something
because of this, this someone thought like this at this time:

b. “something bad happened
    I don’t want this”

c. because of this, this someone felt something bad
like people often feel when they think like this

d. this someone wanted to do something because of this

The explications for both words share components (b) and (c), which means that the experiencer feels (c) as a result of thinking (b). In both cases, something undesirable happens to the experiencer, which the experiencer did not wish for. However, they differ in components (a) and (d), which represent the cause and the expression of the emotion. The use of the verb kanashimu (grieve) suggests that the experiencer is consciously holding a particular thought and involves physical movement, such as crying or crouching.

These explications clarify the differences in meaning between the adjective kanashii (sad) and the verb kanashimu (grieve), as illustrated in the example cited in the Introduction.

(34)  Kanashii toki-nya     kanashimi-nasai

          sad           time-top     grieve-imp.pol

          ‘When you are sad, grieve.’

Unlike the adjective kanashii (sad), the verb kanashimu (grieve) implies that the experiencer actively interprets the situation as negative. This is reflected in the semantic component “this someone wanted to do something” and the emotion is likely to be expressed through action (“this someone wanted to do something because of this”). The lyrics convey the message that, even in the face of negative situations, one should not suppress the resulting emotions but rather recognize and express them.

The proposed template allows us to systematically analyze the semantic structures of Japanese emotion concepts by capturing both the similarities in meaning shared by emotion verbs and the differences in meaning of each emotion concept. However, further investigation is needed to determine its applicability to emotion verbs in other languages, as their meaning is not uniform across languages.¹¹

5.   Discussion

This section re-examines the conceptualization of emotion in Japanese through the lens of Japanese culture. Particularly, it considers whether the relatively infrequent use of emotion verbs, despite their lexical richness, might be shaped by cultural norms that discourage overt emotional expression.

The study of the language of emotion deepens our understanding of human nature and sheds light on the mechanisms of human language and communication (Alba-Juez & Larina 2018, Mackenzie & Alba-Juez 2019). In this context, cross-linguistic comparisons reveal that languages differ in their preferred ways of conceptualizing emotion. English predominantly employs adjectives or pseudo-adjectives to express emotion. It contains relatively few intransitive emotion verbs, such as grieve, rejoice, and pine, which have become less common in modern usage. Other intransitive emotion verbs, such as sulk, fret, fume, or rave, often carry negative or disapproving connotations (Wierzbicka 1995). In contrast, Russian possesses numerous adjective–verb pairs for emotion, and emotions are commonly expressed through verbs rather than adjectives (Wierzbicka 1999: 18). This cross-linguistic variation is further illustrated by Dziwirek and Lewandowska-Tomaszczyk (2010), who compared the lexicalization of six types of emotions—anger, fear, sadness, like, joy, and dislike/disgust—through a corpus study. Their findings show that Polish uses transitive verbs to express emotions more frequently than English.

As shown in Section 3.3, Japanese possesses a richer inventory of emotion verbs than English; however, their usage remains significantly lower than that of emotion adjectives. This discrepancy suggests the presence of linguistic or cultural constraints that discourage the use of emotion verbs, which invite inferences about the expression of emotion. Specifically, the limited use of emotion verbs in Japanese may be attributed to cultural norms that discourage overt emotional expression. This view is supported by the following observation:

(35) In social interaction, Japanese people generally are expected to restrain, if not suppress, the strong or direct expression of emotion. Those who cannot control their emotion are considered to be immature a human beings. Strong expression (verbal or nonverbal) of such negative emotions as anger, disgust, or contempt could embarrass other people. Direct expression of sorrow or fear could cause feelings of insecurity in other people. Expression of even happiness should be controlled so that it does not displease other people. (Honna & Hoffer 1989: 88–90)

Hence, Japanese culture encourages people to refrain from expressing both negative and positive emotions. The following corpus example illustrates how such values are reflected in everyday communicative practices.

(36)    Umarete hajimete   houseki-o   moratte,  ureshikatta.  Shikashi  haha-no         

          born        first.time   jewel-acc  receive    happy.pst       but         mother-gen 

          mae-de             mujakini       yorokobanai   youni     shita. 

          in.front.of         innocently    rejoice.neg     way       did     

          Musume-wa     musume-de   ki-o                 tsukatteiru   noda. 

          daughter-top   daughter-as   mind-acc  use.prog     cop

           ‘I was happy to receive a jewel for the first time in my life. But I tried not to rejoice in front of my mother. As her daughter, I was being considerate.’ (Matsumoto Yuko, Hana no nedoko)

The daughter feels ureshii (happy), upon receiving jewelry for the first time, but refrains from expressing this emotion to avoid upsetting her mother. Consequently, the emotion verb yorokobu (rejoice) is used in its negative form, reflecting a deliberate effort to refrain from the expression of emotion. Thus, Japanese people are “trained from childhood on to suppress external signs of high emotion and to control carefully expressions of pleasure or pain, anger, or joy, love, or hate” (Brosnahan 1990: 99). This highlights how cultural values may suppress situations that are typically conceptualized through emotion verbs.

However, this factor alone does not fully account for the low frequency of emotion verbs. If it were the primary cause, verbs explicitly involving emotional expression would occur less frequently than those that do not. Since this is not the case, cultural constraints should be regarded as only one factor influencing verb usage. Identifying additional factors remains an key direction for future research.

Hence, the relatively infrequent use of emotion verbs in Japanese, despite their lexical availability, may be attributed to culturally embedded values that discourage the overt expression of emotion. Speakers of different languages prefer different grammatical forms for conceptualizing emotion, and these preferences reflect underlying cultural values.

6.   Concluding remarks and remaining issues

This study investigated three differences in the conceptualization of emotion between adjectives and verbs in Japanese: (i) the cause of the emotion, (ii) the object of the emotion, and (iii) the expression of the emotion. It focused on nine adjective–verb pairs, selected from 88 adjectives based on definitions from previous studies. While verbs can be used with nouns that evoke the opposite or unrelated feelings, adjectives do not co-occur with such nouns. This difference suggests that the feelings designated by verbs are caused by the experiencer’s conscious thought, whereas the feelings designated by adjectives are triggered by an external cause. Furthermore, the feelings expressed by some verbs involve physical movement, while others do not.

Based on these differences, this study proposed a semantic template for Japanese emotion verbs, highlighting both the similarities and differences in the meanings of emotion concepts. It has been suggested that the relatively infrequent use of emotion verbs in Japanese, despite their lexical richness, may be linked to cultural norms that place a negative value on emotional expression. These observations imply that the way emotions are conceptualized reflects culturally preferred communication styles, which vary across linguistic communities. Investigating how different languages lexicalize emotions illuminates the relationship between language and emotion, contributing to a deeper understanding of cross-linguistic variation in the use of different grammatical forms for expressing emotion. Further inquiry into this relationship would clarify the factors influencing the frequency and distribution of emotion terms, as well as the cultural motivations underlying these patterns.

Furthermore, additional investigations are needed to determine whether the semantic template proposed in this study applies to other Japanese emotion verbs or to emotion verbs in other languages. As this study employed a method based on the possibility of comparison with formally corresponding adjectives, it is necessary to consider alternative methods to verify the applicability of the template to verbs that do not have corresponding adjectives.

_{¹ Although this phenomenon is known as “person restriction,” scholars have not yet reached a consensus on why it occurs specifically in the domain of emotion (Wang 2013).}

_{² https://clrd.ninjal.ac.jp/bccwj/en/}

_{³ Each sentence remains acceptable even if the subject is changed: the use of a third-person subject in (2a) or a first-person subject in (2b) does not affect acceptability.}

_{⁴ This is consistent with Langacker’s (2008) claim that adjectives profile “atemporal relation,” whereas verbs profile “process.”}

_{⁵ As both the verbs start and stop can be followed by infinitives and gerunds, phrases in both forms were included in the search.}

_{⁶ Semantic primes have been presented in several languages, including Japanese (Asano-Cavanagh & Farese 2015). This study uses the English version of semantic primes to facilitate comparison with the template for English emotion adjectives.}

_{⁷ Even among English emotion adjectives, some require slightly different templates. For example, disgust requires a component related to “the ones that involve part of the body” (Goddard 2018: 76). For more details on the explication of disgust, see Goddard (2014).}

_{⁸ Since the dagger symbol (†) indicates that ureshimu (rejoice) is obsolete, Table 2 lists yorokobu (rejoice), as the corresponding verb for the adjective ureshii (happy) in modern Japanese. Although the suffix -garu can transform adjectives into verbs (e.g., the adjective kowai (scary) into the verb kowagaru (fear)), such verbs are excluded from this analysis because they introduce additional meaning.}

_{⁹ This is presumably because the verb nikumu (hate) is primarily used in written language, and the data in Table 3 are based on a written-language corpus. In the Corpus of Spontaneous Japanese (CSJ), the adjective nikui (hateful) appears 11 times, more than the verb nikumu (hate), which appears six times (NINJAL 2024).}

_{¹⁰ This analysis is consistent with findings on emotion nouns in Japanese. Such nouns can be derived from adjectives or verbs and exhibit the characteristics of their source class. De-verbal emotion nouns can co-occur with nouns that evoke opposite or unrelated feelings, whereas de-adjectival emotion nouns are less likely to do so (Sakaba 2024).}

_{¹¹ It should be noted that this study is limited to emotion verbs where the experiencer is the subject. English has two types of emotion verbs: experiencer-subject verbs, such as like or admire, and experiencer-object verbs, such as please or scare. Because they differ in causal-aspectual semantic type, different templates may be needed for experiencer-object verbs; experiencer-subject verbs are “purely stative,” whereas experiencer-object verbs are “causative” (Croft 1991: 215).}

About the authors

Hiromichi Sakaba

Author for correspondence.
Email: claudia.zbenovich@gmail.com
ORCID iD: 0009-0005-9859-009X

Assistant Professor at the Faculty of Global Communications

References

Supplementary files