The Analysis of Adjunct to Refusal Expressions in a Refusal Discourse: A Contrastive Study between Japanese Native Speakers and Indonesian Japanese Learners

___________________________________________________________________ The study reports analyzing Adjunct Refusal Expressions used by Japanese Native Speaker (JNS) and Indonesian Japanese Learners (IJL) in a refusal discourse. The aims are to see whether both use Adjunct to Refusal Expressions in a refusal interaction and to observe what type of expressions they use when refusing in the interaction, investigating the similarities and differences in their use from the first refusal stage to the stage where the refusee accepts the intent of refusal. The mixed method research involved 60 participants consisting of 30 pairs of JNS female university students and 30 pairs of IJL female university students. The male students were not involved in this study due to the lack of students learning the Japanese language at the university. IJL students represented beginner and intermediate levels (N5, N4, and N3). The data were collected by situating the participants in a role-play based on the setting given while they were being videotaped. The setting is a refusal of a close friend’s request. The data were analyzed by using a semantic formula framework, focusing on Adjunct to Refusal expressions modified from Beebe et al. (1990), Yoshida (2015) and Hayati (2020, 2023). The results showed that both JNS and IJL


INTRODUCTION
In daily communications, refusing has always been considered difficult.This is a complex issue because the refuser, someone who does the act of refusal, has either to directly or indirectly say 'no' to the wishes and expectations of the refusee, someone who receives the act of refusal (Salazar & Jorda, 2009;Bulaeva, 2016;Tuncer, 2016).It contains dispreferred messages (Hayashi, 1996;Yule, 1996;Moaveni, 2014), threatens the desired positive face (Brown & Levinson, 1987;Tanck, 2004), makes the refusee feel unpleasant or uncomfortable due to negative rejection (Moriyama, 1990;Wierzbicka, 1991), and requires the refuser to possess a high level of pragmatic skill (Cohen, 1996;Chang, 2008).
Accordingly, due to its face-threatening nature, refusal constitutes a difficult task in daily communication.Therefore, it is always played out in lengthy processes involving interaction negotiation that takes place more than once until a resolution or final outcome is achieved or at least perceived as so (Wierzbicka, 1991;Gass & Houck, 1999;Ebsworth & Kodama, 2001).
One classification of semantic formula widely used in refusal studies is proposed by Beebe et al. (1990).Their classifications include two semantic formulas (direct and indirect) and adjunct to refusal.Beebe et al. (1990) explain that in indirect refusal, the refusers directly express their refusals, realized by strategies of performative and negative willingness/ ability, while in indirect refusal, the refusers utilize phrases that possess all sorts of mitigating formulas, realized in 11 strategies, including; (1) statement of regret; (2) wish; (3) excuse, reason, explanation; (4) statement of alternative; (5) set condition for future acceptance; (6) promise of future acceptance; (7) statement of principle; (8) statement of philosophy; (9) attempt to dissuade interlocutor; (10) acceptance that functions as refusal; and (11) avoidance.
Further, Beebe et al. (1990) illustrate that adjunct to refusal is the other classification of semantic formula, utilized to supplement the refusals either directly or indirectly, used by the refusers to convey support to the interlocutor and help them save face (see also Eslami, 2010;Morkus, 2014: p.91).In this case, Beebe et al. (1990) classify adjunct to refusal into four, namely: (1) a statement of positive opinion/feeling or agreement, (2) a statement of empathy, (3) pause fillers, and (4) gratitude/ appreciation.
Another classification of the semantic formula was developed by Yoshida (2015), which is modified by Beebe et al. (1990), Fujiwara (2004), Ito (2004), and Yim (2004).The modification lies in the different number of strategies under indirect classification and adjunct to refusal, while the number under direct classification remains the same.In this case, while classification by Beebe et al. (1990) involves 11 strategies under indirect classification, that by Yoshida (2015) includes only seven categories, covering (1) reason, (2) statement of regret, (3) apology, (4) statement of alternative, (5) enquiry about statement of alternative, (6) mention future contact, and (7) set the condition for past or future acceptance.
Concerning the number of strategies of adjunct to refusal, Beebe et al (1990)  Of all those classifications of semantic formula, Adjunct to Refusal expression is the study's primary concern.It focuses on the expressions of Adjunct to Refusal used by Japanese Native Speakers (JNS) and Indonesian Japanese Learners (IJL) within the whole refusal interaction, from the first refusal to the part where the refusee accepts the refusal intent.Therefore, this study highlights the entire interaction process instead of focusing on the linguistic features of refusals seen in one stage of interaction.The purposes are to see whether expressions of Adjunct to Refusal serve as acts of consideration in refusal discourse and to find any similarities and differences in using Adjunct to Refusal expressions between JNS and IJL.
Expression wish is used to show active and positive attitudes, the feeling of being positive by attempting to harmonize with the wish of the requester.A statement of empathy expresses understanding of the requester's situation and approaches the requester.Pause filler/ hesitation functions to state something unclearly, to give a pause filler before stating an opinion/answering.Confusion means that the refuser expresses confusion, having no idea what to do.Information confirmation is done by confirming again/reconfirming more specific details about the interlocutor's matter.A backchannel response is defined as speech usually at the beginning of the refusal discourse to express consideration.Encouraging words indicate that the request was refused, and that the interlocutor is troubled or depressed and then encouraged by the refuser's efforts.Positive expression shows interest and positive feelings about the interlocutor's intentions.Expression of gratitude describes appreciation for the interlocutor's acceptance or understanding of the speaker's refusal intent.Repetition describes the repeated parts of the statements uttered by the requester by lowering the intonation.The request is to request information in detail about the interlocutor's intentions.
Further description of the examples of each type of Adjunct to Refusal expression is summarized in the following table.In line with this, studies in Japanese refusal utterances have long caught the attention of many researchers, and the results have given great insights into the literature of refusal studies.While little research shows that Japanese refusals are likely to be direct (Beebe et al., 1990;Gass & Houck, 1999;Osuka, 2009), especially when the refusees are low in status or power, most Japanese refusal studies reveal that Japanese refuse indirectly almost in all settings (Beebe et al., 1990;Mizutani, 1987in Wierzbicka, 1991;Gass & Houck, 1999;Tsunekawa, 2019), which is realized through different strategies such as apology, reasoning, and postponement (Beebe et al, 1990;Gass & Houck, 1999;Ebsworth & Kodama, 2001).Apart from that, the use of adjunct to refusal is also highly favoured as one strategy to refuse in Japanese context, realized mainly by the use of fillers (Ikeda, 1993;Gass & Houck, 1999;Yamane, 2002;Che, 2006;Nakajima, 2008Nakajima, , 2011;;Katsu, 2015;Yoshida, 2015;Wang, 2016) or by the use of back-channel response and encouraging words (Yim, 2004;Hayati, 2020Hayati, , 2023)).This discourse marker, such as "uuun, aaa, eee" in an utterance, constitutes "asobi kotoba," or "playing words," expressing an intention such as warning, agreement, or refusal (Yim, 2004) and are used to reduce unpleasant feelings, to maintain harmony between the refuser and the refuse and to suggest that the refuser cannot fulfil the request (Ikeda, 1993;Yamane, 2002;Nakajima, 2008Nakajima, , 2011;;Yoshida, 2015;Katsu, 2015;Che, 2016;Wang, 2016).Despite its contribution to Japanese refusal literature, most studies have been centred on refusal utterance.Only a few studies have focused on Japanese refusal discourse; therefore, this study is worth investigating.
As previously stated, this study is carried out to determine whether expressions of Adjunct to Refusal serve as acts of consideration in refusal discourse and to explore any similarities and differences in using Adjunct to Refusal expressions between JNS and IJL.The outcomes are expected to have practical implications for Indonesian Japanese learners, addressing issues in intercultural communication.

METHOD
This research used mixed methods, employing qualitative and quantitative design in seeking the topic under investigation.It involved 60 participants consisting of 30 pairs of JNS female university students and 30 pairs of IJL female university students.The male students were not involved in this study due to the lack of students learning the Japanese language at the university.Furthermore, IJL represents beginner and intermediate levels (N5, N4, and N3).In order to collect the data, the participants were asked to do a role play based on the setting given while being videotaped.The setting is a refusal of a close friend's request.
The data were transcribed by following the framework of Usami (2011) and analyzed by using the framework of Adjunct to Refusal expressions modified from Beebe et al. (1990), Yoshida (2015) and Hayati (2020Hayati ( , 2023)), as illustrated in the Introduction section.
After transcription, the number of adjuncts of Refusal shown in the discourse was calculated using Chi-square so that a significant difference in the use of Adjunct to Refusal expressions between JNS and IJL could be observed.The results were then interpreted and discussed by referring to the framework and the previous studies.

RESULTS AND DISCUSSION
As previously stated, this study is conducted to see whether Adjunct Refusal expressions serve as acts of consideration in refusal discourse and to figure out similarities and differences in the use of Adjunct Refusal expressions between Japanese Native Speakers (JNS) and Indonesian Japanese Learners (IJL).To maintain clarity, both questions are answered separately.

Adjunct to Refusal Expressions Serve as Acts of Consideration in Refusal Discourse
The analysis showed that the use of Adjunct to Refusal was observed in the refusal discourse of JNS and IJL.Its use indicates that both speakers interject some linguistic behaviour as a cushion in the process of exchanging refusals in order to supplement the refusals either directly or indirectly, help them save face and convey support to the interlocutor (Eslami, 2010;Morkus, 2014: p.91).In other words, the study confirmed that participants utilized Adjunct to Refusal expressions as an act of consideration in Refusal discourse.
Of the total 11 Adjunct to Refusal expressions used as the framework in this study, only seven types frequently appear in JNS and IJL refusal discourse.Those include information request, pause filler, back-channel response, confusion, information confirmation, repetition, and wish, as seen in the following figure.The figure above illustrates the types and frequency of use of Adjunct to Refusal expressions by JNS and IJL.In this case, JNS and IJL highly use information requests to supplement the refusals by asking for details on the request and buying time before declining the request.Likely, the choice and use of this discourse marker by both participants were due to considering the close relation between the refuser and refusee so that harmony can be maintained and unpleasant feelings can be reduced (Yim, 2004).
Information request, in this case, serves as an expression of consideration, that the refuser has the intention to consider the request before refusing and that the refuser makes an attempt to give a 'cue' that he or she cannot fulfil the request (Ikeda, 1993;Yamane, 2002;Nakajima, 2008Nakajima, , 2011;;Yoshida, 2015;Katsu, 2015;Che, 2016;Wang, 2016).
In line with this, of the two participants, JNS used Information Request more frequently (79%) than IJL (60%).Even so, the Chi-Square test calculation indicates no significant difference between JNS and IJL in the use of Adjunct to Refusal expressions with χ2(1) = 1.270, p = .2598.It may be attributed to the similar cultural background of the participants.Since both come from high context culture, refusing a request is hard to do and considered negative in some ways (Moriyama, 1990;Wierzbicka, 1991), and therefore, this type of communication is always played out in a long way, involving interaction negotiation that takes place more than once until resolution or final outcome is achieved or at least considered as so (Wierzbicka, 1991;Gass & Houck, 1999;Ebsworth & Kodama, 2001).
Furthermore, the chart also reveals that JNS use high pause filler (50%) compared to IJL (20%).It confirms the studies carried out by Ikeda (1993), Gass and Houck (1999), Yamane (2002), Yim (2004), Che (2006), Nakajima (2008, 2011), Katsu (2015), Yoshida (2015) and Wang (2016), stating that the use of Adjunct to Refusal expression is also highly favoured as one strategy to refuse in the Japanese context, realised mainly by the use of fillers.An utterance such as "uuun, aaa, eee" is used to express the intention of refusal by warning that the refuser will decline the request (Yim, 2004).
Meanwhile, in the case of IJL, the reason of not using filler or hesitation (20%) as often as JNS is likely because of three reasons; (1) their preferences in the use of types of Adjunct to Refusal expressions; (2) the amount of 'closeness' is defined subjectively, giving a hint that the closer the relation the more direct the way they communicate in most settings, including in refusal or conversely, the closer the relation the harder to reject as there are risks that due to the rejection they may end their friendship because rejection revealing dislike messages (Hayashi, 1996;Yule, 1996;Moaveni, 2014) and threatens the maintenance of the harmony between the refuser and refusee (Ikeda, 1993;Yamane, 2002;Nakajima, 2008Nakajima, , 2011;;Yoshida, 2015;Katsu, 2015;Che, 2016;Wang, 2016); and (3) the limited competence of the participants in Japanese language may also become the hindrance in understanding that pause filler/ hesitation can be used as a discourse marker to buy time so that refuser can organize word to refuse the request as also suggested by Finegan et al (1991) and Eslami ( 2010).The analysis also disclosed that in this study, JNS never used confusion (0%) as an Adjunct to Refusal expression, indicating that this expression is not preferable, choosing other types to refuse.Similarly, IJL never used a backchannel response (0%) in their refusal interaction, likely because of either their preferences or their limited knowledge of using this type of utterance as an act of consideration before declining the request.
The research found that JNS used Adjunct to Refusal expressions higher than IJL, as seen in the following chart.

The Use of Adjuncts to Refusal Expressions in Refusal Discourse
The figure shows that JNS use more expressions of Adjunct to Refusal (46,7%) than IJL (33,3%), confirming some previous studies conducted by Ikeda (1993, Gass and Houck (1999), Yamane (2002), Yim (2004), Che (2006), Nakajima (2008, 2011), Katsu (2015), Yoshida (2015) and Wang (2016), stating that Adjunct to Refusal expressions is also highly favoured by Japanese as one of the refusal strategies.This is very likely since the Japanese have a high context culture in which refusal is seen as an unfavourable response, so the adjuncts to refusal expressions appear before the utterance of refusal (Meynard, 1993;Kawate-Mierzejewska, 2002;Yim, 2004) playing a role as a buffer for the refusal in order to soften the impact of the refusal that is about to be made and to give the requester/refusee for negotiation and not being accepted as a sufficient refusal by the requesting party (Qiu, 2002) reducing the risk of a threat to the interlocutor's face and making refusals more acceptable (Zhang, 2013;Hayati, 2020Hayati, , 2023)).
On the contrary, the infrequent number of use of Adjuncts to refusal by IJL can be attributed to their level of learning as they are mostly beginners and intermediate students (N5, N4, N3), so their Japanese language proficiency and sociolinguistic interactional competence are still limited.It has been assumed and becomes one of the weaknesses of this study.Thus, it also implies that both refusers (JNS and IJL) tend not to communicate their refusal intentions during refusal negotiations along with the flow of discourse, and in other words, they tend to avoid straight refusals (Zhang, 2013;Hayati, 2020Hayati, , 2023)).Similarities and Differences in the Use of Adjunct to Refusal Expressions between Japanese Native Speakers (JNS) and Indonesian Japanese Learners (IJL) Concerning similarity, this study found three resemblances between JNS and IJL when using adjunct to refusal expressions in the given setting.First, both have the same preferences for types of adjudication to refusal expressions when attempting to refuse in the interaction.Both used information requests to supplement the refusals either directly or indirectly by asking for more details or information on the request in order to reduce unpleasant feelings as well as to maintain harmony (Ikeda, 1993;Yamane, 2002;Nakajima, 2008Nakajima, , 2011;;Yoshida, 2015;Katsu, 2015;Che, 2016;Wang, 2016).
Second, most JNS and IJL use information request expressions in their first attempt to refuse.Depending on the length of the process of refusal discourse, this type of expression was again applied, which is usually an indication of recurrent efforts of refusal strategies employment from the refuser and responses from the refusee (Gass&Houck, 1999) and are likely to distinctively verbalize, organized appropriately by paying attention to the situation (Finegan et al., 1991;Eslami, 2010).
Meanwhile, the differences in the use of Adjunct to Refusal expressions by JNS and IJL were realized in their preferences towards certain types as they were involved in the refusal discourse.For example, in the case of JNS, the participants in the study preferred not to use confusion (0%), yet IJL used this expression relatively high (30%).This high frequency of use in IJL can be attributed to the influence of cultural background on the participants' mother tongue.Most IJL participants are Sundanese, and in its culture, confusion expression is highly used to supplement the refusals, as also stated in a research carried out by Hayati (2023) explaining that compared to JNS, expression of confusion was highly used by Sundanese Native Speakers as Adjunct to Refusal expression in refusal interaction.

CONCLUSION
The study reports on using Adjunct to Refusal expressions in a refusal discourse between Japanese Native Speakers (JNS) and Indonesian Japanese Learners (IJL).It is carried out to determine whether Adjunct Refusal expressions serve as acts of consideration in refusal discourse and to investigate similarities or differences in the use of Adjunct Refusal expressions between Japanese Native Speakers (JNS) and Indonesian Japanese Learners (IJL).The participants are asked to do a role play based on a setting where they are close friends, in which one acts as a refuser to a request.The results showed that using adjucts to express refusal was observed in refusal discourse.
Both JNS and IJL interject some linguistic behaviour as a cushion in the process of exchanging refusal.Refusers tend not to communicate their refusal intentions during refusal negotiations along with the flow of discourse and are highly likely to avoid straight refusals.Furthermore, Japanese Native Speakers are more frequently Adjunct to Refusal Expressions than Indonesian Japanese Learners.The highest type of Adjunct to Refusal expression both use is Information Request.The difference lies in the order of use of the expression apart from the Information Request.
In JNS, refusers use hesitation and backchannel response, respectively, while in IJL, refusers use confusion and hesitation.Since the participants of IJL involve beginner and intermediate level respondents (N5, N4, N3), the results of the study are likely to be affected by their low Japanese language proficiency and sociolinguistic interactional competence and therefore, it is suggested that in the future study, the participants involved are Japanese language learners with higher levels such as N2 and N1.

Figure 1 . 1
Figure 1.1 The Use of Adjunct to Refusal Expressions by JNS and IJL