Complaint Discourse Features in Modality Forms A Comparative Analysis of Japanese and Indonesian Internet Review

This paper describes a comparative study of complaint discourse in Japanese and Indonesian focusing on modality expression. 500 items of each language were collected from TripAdvisor and usage of modality were quantitatively and qualitatively analyzed. As a result, modality forms have often been reported to be used to show consideration for the other person, such as avoidance of judgments and euphemisms, roundabout expressions or consideration ( hairyo ) for the other person are expressed. It was confirmed in complaint discourse, that there are cases in which have no consideration ( hairyo ) for the other person at all, rather than in a direct way of speaking. Both Japanese and Indonesian modality functioned effectively for expressing complaint. It was found that in “obvious complaint” the modality emphasizes the hotel’s faults and forces them to work appropriately, and in “implicit complaint”, the modality can express complaint with slightly reducing the burden on the hotel side by emphasizing the self-pay. The subject is used to determine whether an utterance expressing Complaint is “explicit/direct” or “implicit/indirect” in terms of the discourse level. In other words, the degree of FTA (Face Threat Act) differs depending on whether the subject is the writer (the guest) or the reader (the hotel). When the subject of “an act” is the reader (hotel side), the modality form functions to emphasize the reader’s fault (wrongness) or to force the reader (hotel side) to act. On the other hand, when the subject of “an act” was the writer, the function was to emphasize self-imposed burden or to understate the fault/burden of the reader (the hotel side).


INTRODUCTION
Conventional research on expressing complaints has focused on the direct expression of complaints as one of the verbal behaviors in the fields of sociolinguistics and Japanese language education , and discourse studies have been conducted by comparing Japanese learners and Japanese discourse by learners (Hatsushikano, Kumatoridani, & Fujimori, 1996;Lee, 2006, etc.).
Expressing Complaint is an act that threatens the other person's position, and it is a difficult act that can have a large impact on human relationships depending on how people use the words and how they talk (Olshtain & Weinbach, 1993;Hatsushikano, Kumatoridani, & Fujimori, 1996, Yamaoka , 2004 ).Also , as pointed out by Wongsamin (2016), In (1), by stating the proposition itself in the categorical form, we can feel the nuances of judging the other party.It is thought that (2) and (3) avoid making a conclusion or obliquely point out the other party's actions that have caused a disadvantage by expressing conjecture and possibility judgment.Sometimes, "maybe" has the function of avoiding the risk of FTA (Face Threatening Act) in which one's own utterance threatens the negative face of the other party (Yamaoka, 2016).
From the above, it is considered that the mode of expression of Complaint is closely related to the final expression (modality).However, the research on complaint expressions so far has focused on pragmatic analysis based on discourse markers, and there has not been much analysis of speech function in relation to grammatical forms.In particular, the modality forms are important in expressing complaints such as used based on what kind of function, how the usage differs in different languages, and what intentions the speaker has in using the modality form, are not sufficiently analyzed.
Therefore, in this study, while relying on the results of previous complaints research, we will conduct a comparative analysis of internet review data of complaints expressed by Japanese and Indonesian speakers, with a particular focus on the use of modalities.By doing so, we will clarify the characteristics of Complaint expressed by Japanese and Indonesian speakers that have not been clarified so far.

RESEARCH FRAMEWORK
Research on Expressing Complaints Ishizuka (2014) conducted a survey analysis of complaint expressions of customer service among Japanese native speakers and Korean learners of Japanese.According to Ishizuka, among native Japanese speakers, strategies of 'proposition' and explicit 'request for improvement' were often seen.Specific examples include, 「確かにそのような掲 示があったんでしょうか」(Tashika ni sono you na keiji ga attandeshouka) 'Is there really such a notice?' and 「セール品であっても人が不快を感じる商 品なら売るべきではないと思います」(Seeru-hin de attemo hito ga fukai o kanjiru shouhin nara urubeki dewa nai to omoimasu) 'Even if it's a sale item, I don't think you should sell it if it makes people feel uncomfortable'.Wongsamin (2016) examines expressions of Complaint and responses to them through roleplays targeting Japanese and Thai native speakers.As a result, it has been clarified that Japanese native speakers (JNS) use strategies that seek explanations of causes and reasons, and objectively emphasize the consequences of situations.Specific examples include requests for explanations of reasons and the provision of general information, such as 「え↑何で辞めちゃうの？」(E↑ nande yamechau no?) 'What?↑Why are you quitting?' or 「 だ け ど 明 日 も う 始 ま っ ち ゃ う で し ょ 」 (Dakedo ashita mou hajimachatta yo!) 'But it's starting tomorrow isn't it?'.On the other hand, Thai native speakers (TNS) emphasized their own disadvantages, and in addition, it was found that sarcasm and jokes were used as a characteristic strategy of TNS.As a specific example, the expressions such as 「せっかく紹介してあげたの に，私の面目が潰れちゃうじゃない」(Sekkaku The overall tendency of Complaint expression strategies among native Japanese speakers is 'do not express Complaint' or even if they express Complaint, 'request for reason/explanation', 'express Complaint in a roundabout way', and use 'euphemistic expressions' (Choi, 2009;Lee, 2006;Jeong, 2005).
Puksi (2017) studied complaint expressions using an approach closest to this study.Puksi (2017) analyzed complaint speech act on an accommodation reservation site in Indonesia, and focused on review postings.Puksi (2017) examined complaint strategies and found that review posting is also related to actions that infringe on the other person's face, and the most commonly used complaint strategy is 'direct accusation', and 'annoyance'.In other words, it can be said that Indonesians prefer to use explicit and direct strategies and directly complain without considering the face of the reader (hotel side).
However, Puksi (2017) leaves some limitations.First, 160 cases of data are considered insufficient to understand complaints overall.In Puksi's (2017) study, the most fundamental problem is that there is no clear standard for identifying which of the reviews posted is a sentence with 'complaint'.Since there are no clear criteria, the subject of analysis may be subjective and prejudiced.To fill in this gap, as mentioned in Method section in this paper, we will not observe 'sentence' but 'discourse' of complaint as the object of analysis.
Based on the research on complaints, it is found that there are two aspects to the way complaint is expressed, and the criteria of 'explicit/direct' or 'implicit/indirect' can be different depends on the country and language.However, it is not clear what kind of clues are used to determine whether complaints are 'explicit/direct' or 'implicit/indirect'.Therefore, in this paper, we are also examining the clues necessary for judging whether Complaint utterances are 'explicit/direct' or 'implicit/indirect' at the discourse level.

Relation Between Expression of Complaint and Expression of Consideration
In recent years, complaint expression studies have pointed out the importance of focusing on modality-type behavior (Yamaoka, Makihara, & Ono, 2010;Yamaoka, 2016).Modality represents the speaker 's attitude toward the proposition and the listener , and is one of the ways to understand the utterance as an expression of complaint .Yamaoka , Makihara , and Ono (2010 ) and Yamaoka ( 2016 ) conducted research on expressions of Complaint within the study of considerate expressions.He states that the mental attitude expressed and the communicative attitude expressed by the final particle include the consideration to maintain the interpersonal relationship as good as possible .Therefore , the modality format is used to avoid the risk of FTA (Face Threatening Act ) by intimidating the negative face of the interlocutor , to convey one's expectations without blaming the interlocutor , and to convey empathy.Yamaoka, Makihara, and Ono (2010) point out that expressing Complaint constitutes an FTA (Face Threatening Act), which may damage human relations and is related to Leech 's (1983 ) politeness principle.Leech (1983) has the following principles regarding evaluation of others.Since expressing complaint is 'blaming others', it should be avoided as much as possible or express it passively (Yamaoka, Makihara, & Ono, 2010, p. 183).In addition, Yamaoka, Makihara, and Ono described that the mental attitudes expressed in modality including end sentence particles are important, and often include considerations which help to improve the interpersonal relationship.Yamaoka, Makihara, and Ono (2010) and Yamaoka (2016) discuss that 'kamo shirenai', 'temo ii' and 'mono (end sentence particles)' are often used to avoid the risk of FTA (Face Threatening Act) that threatens the negative face of the interlocutor, to convey one's expectations without blaming the other party, and to convey empathy.
According to Yamaoka, Makihara, and Ono (2010) and Yamaoka (2016), it is possible that modalities are actively used in complaint discourse.To confirm this prediction, this study decided to analyze the relationship between complaint discourse and modality form deeper.By analyzing the modalities, this study expected to clarify the characteristics of complaint discourse, and the rule of pragmatic modalities usage in general.e-ISSN 2528-5548 | p-ISSN 2776-4478

The Framework of The Analysis
There are representative studies on Japanese modality theory, including Nitta (1991), Miyazaki, Adachi, Noda, andTakanashi, (2002), andJapanese Descriptive Grammar Research Group (2003).On the other hand, Alwi (1992) is the only study representing Indonesian modality theory.However, the modality theory in Japanese and Indonesian have very different frameworks, and it is expected that it will be difficult to analyze if these modality theories are adopted.
Hence, this research used the framework of Palmer (2001)

RESEARCH METHOD Data Collection
The data was collected from reviews posted on the Japanese version of TripAdvisor and the Indonesian version of TripAdvisor.A total of 1,000 reviews were collected, including 500 reviews in Indonesian and 500 reviews in Japanese.Data were collected by the following procedure.We extracted reviews written between 2015 and 2019.For Japanese data, the targeted reviews are those written between 100 to 1,000 characters, and for Indonesian, the targeted reviews are those written between 30 to 800 words.The target hotels are hotels in the metropolitan area in both Japan and Indonesia.
In addition, among the 5 levels of hotel evaluation ("very good," "good," "average," "bad," and "very bad"), only reviews with "very bad" and "bad" ratings were included in this study.We defined it as 'Complaint discourse', and analyzed the modality forms that appear in the whole complaint discourse.On the other hand, the reviews that evaluated the hotel as "very good" or "good" was analyzed as 'satisfaction discourse'.
In this study, we compared 'Complaint discourse' and 'satisfied discourse' to investigate the emergence of modalities in Complaint discourse.The extraction procedure for satisfying discourse was the same as for Complaint discourse, and 250 data were collected.The total number of words in Complaint discourse in Japanese was 97,045 words, and the total number of words in satisfied discourse was 35,030 words.In contrast, 63,539 Indonesian Complaint discourses and 23,448 satisfying discourses were collected.

Data Analysis
This research consists of two stages: quantitative analysis and qualitative analysis.First, we clarify the usage trends of modalities appearing in wordof-mouth submissions by quantitative analysis.Based on the results of the quantitative analysis, we clarify the functions and usage context of the modality form through quantitative and qualitative analysis.For the quantitative analysis, we use the free software "KH Coder" for quantitative text analysis.KH Coder was developed by Koichi Higuchi (2014) of Ritsumeikan University, and is software that can extract words, search documents, perform morphological analysis, and set and aggregate search conditions (coding rules).Since KH Coder does not read the modality format, it is necessary to specify the vocabulary (modality format) to be extracted in advance using a function called "word selection".Since most modality forms are compound forms, there are many exceptions to selection by part of speech.
Therefore, in this research, the modality form to be extracted is specified in advance using a function called "Forced Extraction Word Specification", and forced extraction is performed.In addition, based on the results extracted by KH Coder, a function called KWIC concordance is used to confirm and consider the usage trends of modality format in posting reviews.Since most modality forms are compound forms, there are many exceptions to selection by part of speech.Therefore, in this study, we registered the modality complex form to be extracted in advance by using the function of 'Forced Extraction Word Specification' and forced extraction.

FINDINGS AND DISCUSSION
According to Matsuki (1992), the point of gaze (the object seen) is mainly placed on the actor, while the person on whom the point of view (the place to see) is placed indicated by the verb of the sentence related to the point of view.In line with this point, this study focuses on the point of gaze and clarifies the function of modality forms in discourse of Complaint.
From this section onwards, the 'gazing point' is referred to as the 'subject' and refers to the actor.e-ISSN 2528-5548 | p-ISSN 2776-4478 Depending on the person who is the subject, the manner in which Complaint expressed is different.Another important factor in the characteristics of Complaint discourse in Japanese and Indonesian is the 'feasibility of event'.This point will be explained through a case analysis.

Complaint Discourse in Japanese
Table 1 shows the appearance frequency of modalities in Complaint discourse and satisfied discourse in Japanese.A chi-square test of independence was performed using a crosstabulation table.From Table 1, when the p-value level is <0.05 (significant difference), the Epistemic modality, evidentiary modality, and Deontic modality showed significant differences among the four modalities.Only the communicative modality was non-significant.From this, it can be said that the Epistemic modality, the evidentiary modality, and the binding modality are actively used in Complaint discourse.
This will focus only on the evidentiary modality and the binding modality in Complaint discourse in Japanese.For the list of modality forms, refer to Japanese Descriptive Grammar Research Group (2003) and discuss qualitatively in "Evidential modality" section.
According to Morita (1990) and Kekidze (2003), '(shi) sou' is an expression that avoids assertion against things that can be affirmed, or an expression that is uttered out of consideration and consideration for the other party.However, the '(shi) sou' in Complaint discourse has a nuance of actively negatively evaluating the reader (the hotel side), unlike the meaning that has been reported so far as showing consideration.
In Japanese Descriptive Grammar Study Group (2003, p.173), it is mentioned that '(shi) sou' expresses the subject's disposition and internal state being observed as an external appearance.and being involved in 'already-realized events', '(shi) sou' came to have the function of 'positive negative evaluation'.In the case of "positive negative evaluation", the writer (guest) uses '(shi) sou' to express a positive negative evaluation of the reader's (hotel's) response and facial expressions to emphasize the bad points.
This will be described below with specific examples.
(1) This time as well, I made a reservation from the official website of 〇〇〇 (hotel name), and when I made a reservation for "KING PREMIUM DELUXE," I received a reply email saying, "KING PREMIUM DELUXE, non-smoking confirmed."I thought it was secured, but when I checked in, I was told in a natural way, "The non-smoking room is full and it will be a smoking room!" "It's just a request, not a promise, so the explanation below doesn't make sense."When I asked again, they said, "There is a non-smoking room for twin rooms." In data (1), the subject of 'mendōkusai (troublesome)' is the reader (the hotel), and the state of 'mendōkusai (troublesome)' is the internal state of the reader (the hotel), which can be observed from the expression of the reader (the hotel).Here, the writer (guest) is thought to have emphasized the poor response and facial expression of the reader (the hotel) by adding "sō".In other words, '(shi) sou' fulfilled the function of 'active negative evaluation'.From this, there is a possibility that the face of the reader (hotel side) is threatened, and it seems that Leech (1983)'s approval maxim was violated.
Another factor is the 'feasibility of the event', and '(shi) sou' means 'an event that has already happened'.In other words, there is a nuance that the writer (customer) focuses on the finished situation and strongly points out the other party.In fact, except for '(shi) sou', it is possible to simply write "Akiraka ni mendoukusai kanjide...(It's obviously troublesome...)", but the writer (customer) intends '(shi) sou'.It is thought that by using it in a generic way, it is easy to take on the nuance that emphasizes the undesirable situation, such as 'mendoukusasouna kanjide...(it seems to be troublesome)'.Similar examples is as follows.
( After that, go up to the 28th floor and head to the restaurant.
Confirm the payment we were talking about earlier.
They had asked for it to be handled in a way that was a little time-consuming, but when it came to the day, they would like to avoid that method if possible, as it would be troublesome for the customer if they (the hotel) made a mistake.But it was the person in charge who proposed the method in advance.Why do you say you can't do it on the day of the day?We split the bill, but when we tried to pay at different times, he said in a languid tone, "Is there anyone who hasn't paid yet?" (Omitted) The state of (2) 'kidarui (drowsiness)' is the internal state of the reader (the hotel side) and can be observed from the reader's (the hotel side) facial expression.Here, the writer (guest) is thought to use '(shi) sou' to emphasize the bad point (mistake) of the reader (hotel side).In reality, except for '(shi) sou', it is possible to simply write 'kidarui toon de iwareta (said in a languid tone)'.It is thought that by using it in a generic way, it is easy to take on the nuance that emphasizes an undesirable situation, such as 'kidarui toon de iwareta (said in a languid tone)'.
In Complaint discourse, 'nakereba naranai/naito ikenai (must)' expresses that it is unacceptable that the situation does not happen, that it is indispensable, but when the subject is the 'writer' and the 'actual event' already happened, this fulfill the function of "clarifying the responsibility of others"."Clarification of the responsibility of others" means that the reader (hotel side) is not responsible for a situation where the writer (guest) is doing unnecessary acts.It was confirmed that 14 cases of 'nakereba naranai/naito ikenai (must)' fulfilled this function.A specific example is the following (3).
( All in all it was the worst stay ever.
At check-in, I was made to wait for a long time at the front desk, and while looking sideways at people checking in one after another, I kept standing without any explanation from the front desk.Finally, when I got impatient and asked what the problem was with the person in charge at the front desk, they said they couldn't find the reservation.I gave the confirmation number and finally check-in proceeded.
If I didn't ask him anything, how long would he have to stand after that?(Omitted) It is the worst hotel where it is low level and does not know where to fix anything.
In (3), the writer (guest) actually did the act of 'tatte iru (standing)', even though it was not necessary to do so.The writer (guest) is the actor of 'tatte inakereba naranai (I must stand)', and the writer (guest) gives the impression that he/she is a victim and that the reader (hotel side) made him/her wait for a long time and ask for the hotel responsibility.
If we exclude 'nakereba naranai (must)', the phrase 'ano ato, dorekurai tatteitanodeshou (how long have I have been standing after that?)' does not have the intention of questioning the reader (the hotel side), and is simply an utterance such as 'selfconfirmation' or 'internal speech'.In other words, it is thought that the writer (guest) of (3) chose a strategy of indirectly expressing Complaint using 'nakereba naranai (must)'.In addition, (3) is related to 'events that have already happened', so the writer (guest) emphasized the bad points of the reader (hotel side).More direct terms also used as seen in example (4).In ( 4), the writer (guest) is trying to clarify where the responsibility lies with the reader (hotel side), which is common to (3).However, in (4), the use of 'bekida (should)' allows the writer (guest) to point out the bad points (mistakes) of the reader (hotel side) in a straightforward manner, and the reader (hotel side) takes responsibility.The writer (guest) explicitly expressed his Complaint with the reader (the hotel) which made him/her waited.On the other hand, the subject of 'tatte iru (standing)' in (3) is the writer (guest), and it is thought that he expressed his Complaint more obliquely and indirectly than in (4).In the following (5), as in (3), the writer (guest) expresses his burden grandly.
( I've always wanted to stay at ○○○ (hotel name).The staff at ○○○ (hotel name) was at a level where I had to call out to them, and I had the impression that they were picking out customers and responding to them, so I had a bad feeling from the time I checked in.
The premonition is true with the passage of time.
(Omitted) The staff's processing ability is at the level of a business hotel, such as lack of explanation when providing room service, and not being able to communicate well after repeated contact when reserving a restaurant.(Omitted) I think the staff's service still has room for improvement.But I don't expect much.
In data (5), since the act of 'koe o kakeru (calling out)' was not carried out by the hotel, the writer (guest) had to carry out the act of 'koe o kakeru (calling out)' to the hotel staff.In other words, the act of 'calling out' should be the responsibility of the reader (hotel side), but in reality this is not the case, and the writer (guest) must tried to make the reader (the hotel side) who had to call him/her aware of where the responsibility lies, while using it to express his burden clearly.In (3) and (5) above, the writer (guest) expresses his own burden more than (4), so the risk of face infringement of the reader (hotel side) can be avoided.

Complaint Discourse in Indonesian
This section examines the emergence of modality forms in Complaint discourse in Indonesian.Table 2 summarizes the appearance of modalities in Complaint discourse and satisfied discourse in Indonesian.
From the data presented in Table 2, the Epistemic modality showed a p-value of 0.0002, indicating a significant difference.Another significant difference was found in Deontic modalities.From this data, it can be said that the binding modality was actively used in Complaint discourse.Next, the analysis will be focusing only on Epistemic modalities and binding modalities.As the Indonesian modality formats listed based on Alwi's (1992).
According to Alwi (1992), 'Semoga/Mudahmudahan' is basically used as an adverb, and its basic meaning is to describe an uncontrolled situation.It is pointed out that it represents It is also said to express the speaker's wishes and expectations toward God.In Indonesia, which has a strong relationship with religion, 'Semoga/Mudah-mudahan' is often used in daily conversation as a word that expresses 'prayer' or 'hope' to God.
In Complaint discourse, the writer (customer) can use expressions with a high degree of assertion, such as imperative and request, but by using 'wish/hope', they are passively making efforts to convey they means to the reader.Therefore, in this paper, we refer to this as a "passive demand for action".From the data, there are 22 use of 'Semoga/Mudah-mudahan' which found to perform this function.The characteristic of 'Semoga/Mudah-mudahan' use is shown in a specific example as (6) below.
Tp sayangnya untuk 〇 yg satu ini jauh dari harapan.Apalagi perjalanan dari 〇hotel lanjut ke 〇hotel yg berkualitas, ujungnya 〇 yg levelnya jauh beda padahal harga yg hampir sama.Mudah-mudahan bisa lebih ditingkatkan (www.tripadvisor.co.id) The hotel name with the appendage 〇, of course, is not just a name, but also a description of the quality of service quality.But unfortunately for 〇 this one is far from expectations.
Moreover, the trip from 〇hotel to 〇 a quality hotel, in the end 〇 with a much different level even though the price is almost the same.
Hopefully can be more be improved The sentence (6), 'Mudah-mudahan bisa lebih ditingkatkan (I hope it can be improved)', is expressed as if the writer (guest) asked 'God', and hoped that the hotel would respond by God's power.However, 'Mudah-mudahan' in (6) no longer actually expressed a 'wish' to God.The subject of "Improve" is the reader (the hotel side), and they demand the realization of the act of "Improvement".
However, even if the subject of "Improve" is the reader (hotel side), by using "wish/hope", the reader reached passively and refrain from asserting.Therefore, it is thought that the risk of face infringement on the part of the reader (hotel side) can be avoided because the approach to the other party is not strong and the criticism of others is minimized.From the perspective of 'feasibility of events', 'Mudah-mudahan' means 'unrealized event'.In other words, the writer (guest) focused on the situation related to what the hotel had to do in the future.
From the above (6), 'mudah-mudahan' does not originally have the function of requesting action, but in Complaint discourse, it is thought that it acquired the function of requesting action from 'prayer'.Even though the writer (guest) uses the 'prayer/wish' to direct a certain action (improvement), it is up to the reader (the hotel side) e- ISSN 2528-5548 | p-ISSN 2776-4478 to accept or reject the action.The function of such a "passive action request" can be better understood by comparing it with (7) below, which makes an action request more explicitly.In (7), the writer (customer) made the reader perform the act of 'tlp/telpon (calling)' using the communicative modality 'dong'.According to Kridalaksana (2011 ), 'dong ' has the function of pointing out the mistake on the part of the addressee (directly pointing out the reader 's mistake) like in (7).In (7), the writer (guest) uses ' dong (yo)' to directly point out and criticize the reader's (hotel's) mistakes.
"dong" is more straight forward than a modality that expresses a wish/hope, such as "Semoga/mudah-mudahan".In other words, there is a high possibility of infringing on the face of the reader (hotel side).On the other hand, 'Semoga/mudah-mudahan' has the function of indirectly making the reader (the hotel side) act, so the utterance was relaxed and the expression became a roundabout expression.
Looking back, in the Complaint discourse in Japanese, there is not a single example of the type that compels the reader to act, and it is said that the reader (the hotel side) is made aware of where responsibility lies.However, in the Complaint discourse in Indonesian, unlike Japanese, 'Harus' was found to force the reader to act.
In Complaint discourse in Indonesian, the writer (customer) strongly and positively appealed to the reader by using expressions with a high degree of assertion such as 'Harus (must)'.Accordingly, in this paper, we have referred to it as a "positive demand for action".Twenty cases of 'Harus (must)' with the function of ''positive action demands'' were confirmed, as example (8) below.In (8), "Harus (must)" is used.This created the impression that the writer (guest) was giving strong orders or pointing out to the reader (hotel).The writer (guest) used 'Harus (must)' to ask the reader (hotel side) to act in a straightforward manner, 'doing a job properly like a professional'.'Harus' expresses 'an unrealized event', like 'Semoga/mudah-mudahan', which is good.In other words, it was found that the writers (guests) paid more attention to the improvement points on the hotel side.
"Harus" was used to indicate that the writer (guest) was in a higher position than the reader (the hotel).In terms of social status, the position of 'customer' is considered to be higher than the position of 'clerk', and has more power.The reason e-ISSN 2528-5548 | p-ISSN 2776-4478 why the function of 'positive demand for action' appears only in Indonesian is that Indonesian society is very conscious of the fact that 'guest' is a person of high social status and power.

Discussion
The results and analysis results are described above.It has been reported that conventional modality forms are often used to show consideration for others, such as avoiding assertions and euphemistic expressions (Yamaoka, Makihara, & Ono, 2010;Yamaoka, 2016).It was also confirmed that the discourse does not simply express consideration for the other person in a roundabout way, but is rather direct and does not include consideration for the other person at all.Details are summarized in Tables 7 and 8. C. The error/burden on the part of the reader (the hotel side) has been reduced.

✓ ✓
From Table 7, the modalities of Complaint discourse in Japanese and Indonesian had two aspects: 'explicit complaint' and 'implicit complaint'.In the case of 'explicit complaint', the modality format emphasizes the bad point (mistake) of the reader (hotel side) or forces the reader (hotel side) to act.On the other hand, in the case of 'implicit complaint', the modality format emphasizes self-payment, cares for the reader (hotel side), relaxes the assertion, and moderates the fault/burden of the reader (hotel side).
In this regard, the results are somewhat different from those of previous complaint studies.
Japanese native speakers (JNS) use a strategy of objectively emphasizing the outcome of the situation when expressing their complaint (Wongsamin, 2016), not expressing complaint, or even expressing complaint, 'indirect complaint' and it is mentioned that there is a high tendency to use the 'euphemistic expression' strategy, but by observing the modalities of Complaint discourse, it has two aspects: 'explicit complaint ' and 'implicit complaint'.This research is a further development of how modality forms other than 'kamoshirenai (maybe)', 'temo ii' and 'end sentence particles (mono)' are related to complaint as found by Yamaoka, Makihara, and Ono (2010) and Yamaoka (2016).However, unlike the results reported by Yamaoka, Makihara, and Ono (2010) and Yamaoka (2016), the modality format in complaint discourse does not only include consideration for maintaining good interpersonal relationships with the other party.In complaint discourse, when the subject of "certain action" is the reader (hotel side), the modality format emphasizes the reader's bad point (wrong) or forces the reader (hotel side) to act fulfilled its function.It may infringe on the reader's face because it greatly expresses criticism to others.On the other hand, when the subject of 'a certain act' is the writer, it has the function of emphasizing self-burden and moderating the fault/burden of the reader (the hotel side).By doing so, it is thought that the risk of face infringement on the part of the reader (hotel side) can be avoided.
In this way, we observed characteristics that differed from dialogue situations that had been mainly dealt with in previous research on expressions of complaint.The approach of clarifying the characteristics of complaint expressions in Japanese and Indonesian by focusing on the modality form is also valid, and it is expected that new knowledge about linguistic differences in complaint expressions will be obtained.
Finally, the similarities and differences between Japanese and Indonesian are as summarized in Table 8.
Perspective: A Comparative Analysis of Japanese and Indonesian Internet Review 70 | P a g e e-ISSN 2528-5548 | p-ISSN 2776-4478 Perspective: A Comparative Analysis of Japanese and Indonesian Internet Review 72 | P a g e e-ISSN 2528-5548 | p-ISSN 2776-4478

Andhini Pratami Putri Rustandi,
Complaint Discourse Features in Modality Forms Perspective: A Comparative Analysis of Japanese and Indonesian Internet Review e-ISSN 2528-5548 | p-ISSN 2776-4478 shoukai shite ageta noni, watashi no menboku ga tsuburechau janai) 'I've taken the trouble to introduce you, but you're going to ruin my face' were used.

Table 1 :
Appearance frequency of modalities in Complaint and satisfied discourse in Japanese.

Table 2 :
Appearance frequency of Complaint discourse and satisfied discourse in Indonesian.

Table 7 :
Characteristics of expressions of Complaint seen from modalities.