An Investigation into the Speech Act of Compliment Response in Persian

The purpose of this study is to investigate the compliment responses in Persian language. Drawing on Herbert’s (1986), Mile (1994), Rose and Kwai-Fun (2001), and also the data of the current study, 18 types of compliment responses were identified. Based on different compliment topics [such as appearance, ability, possession and nationality], the researcher along with 12 others as assistants who were instructed how to give compliments got engaged in conversations with different people to compliment them. They complimented people from various professions, contexts, cities and educational levels. The majority [67.6%] of our corpus of 756 naturally occurring compliment responses fell into the main category of “agreement”. However, the absence of the compliment response “no-acknowledgement”, the high rate of two or three part appreciation token along with a significant portion of “ making offer” attest to the culturallyspecific Iranian features such as modesty and Ta’arof. Additionally, no significant differences were found between males and females in terms of compliment response types, or the degree of modesty or Ta’arof. Nonetheless, there existed significant differences with regard to topics of compliment in a sense that most of the compliments given on nationality (89%) were accepted while in the case of physical and mental domain, it was not that much high. Like other studies in pragmatic area, this study also indicates that compliment response is cross-naturally varied, different, and hence warrants more attention in instruction and material development.

Since the emergence of Hyme's (1971) communicative competence, pragmatic features of L2 have come to the fore.In recent years, this aspect of language has been given due attention both in different models of communicative competence for testing purposes ( Backman, 1990;Canale and Swain, 1980) and consequently, in teaching practices in the form of implicit or explicit instruction ( Bardovi-Harlig & Griffin, 2005;Koike & Pearson, 2005;Martinez-Flor & Fukuya, 2005;Rose, 2005& Soler, 2005).Kasper and Rose (2001) pointed out that interlanguage pragmatics is concerned with the "study of how people accomplish their goals and attend to interpersonal relationships while using language" (P.2).They further point out that much of what learners need for development of pragmatic competence already exists in their L1 pragmatic repertoire.In other words, positive transfer from L1 to L2 would facilitate learner's task in acquiring the socio-pragmatic knowledge.Indeed, the key insight gained from comparative research program which marked the inception of inter-language pragmatics (Jeon &Kaya, 2006) is that existing pragmatic knowledge plays an important role in developing L2 pragmatics, and that learners can purposefully build on it (Bialystok, 1993).At the same time, many researchers (Wolfson, 1982a;Rose & Kwai-fun, 200 & Yoshimi, 2001) who are concerned with instruction of L2 pragmatic features found out that there are much variation with regard to both pragmalinguistic and sociopragmatic aspects of L2 and L1 so that learners' attention has to be drawn to these differences.
Given the key role of the existing pragmatic knowledge both in terms of positive transfer (Bialystok, 1993) and variation between the two languages and how this variation can be used to help students notice (Schmidt, 1993) the differences between two languages, this study aims at investigating the compliment responses as a speech act in Persian language among Iranians.Although there exist some studies concerning compliments and compliment responses in English and some other languages such as Japanese, Korean and Arabic, to our knowledge, no study has investigated compliment responses in Persian language using naturally occurring data.This motivated us to carry out this study by analyzing a corpus of naturally occurring conversations with compliment responses as our main concern.Our hope is that the findings of this study can have a contribution to the interlanguage pragmatic competence of Iranian learners of English and those who want to learn Persian as their second or foreign language.

Pragmatics and interlanguage pragmatics
Pragmatics which was initially proposed in Morris ' (1971) semiotics is mainly concerned with the study of the relationship between the signs and their interpreters or between the language and its users in social contexts (Schiffrin, 1994&Yule, 1996).By the same token, interlanguage pragmatics is mainly concerned with the study of non-native speakers' production and comprehension of the pragmatic features of the target language including implicatures, politeness strategies, presuppositions and speech acts.Additionally, ILP is concerned with how foreign or second language learners acquire the pragmatic competence in another language (Kasper, 1992).
In recent years, there has been a growing interest in studies dealing with interlanguage pragmatics.As mentioned before, learners' pragmatic knowledge of their first language and culture plays a critical role in acquiring the pragmatic features of the target language (Kasper, 1992;Wolfson, 1982, 1990& Dash, 2004).The aforementioned researchers ascribed either a positive role to the pragmatic transfer of L1 by stating that the social and cultural norms between two languages are similar, or a negative role by stating that in many cases, non-native speakers of a language cannot communicate appropriately due to the unwarranted generalizations from their first language to the second language.Therefore, in either case, a strong need is felt for investigating the pragmatic features of the two languages so that learners become cognizant about the similarities and differences of two languages.In the next section, a number of studies which have investigated the speech act of compliment in English and some other languages are presented.

Previous studies on speech act of compliment
According to many researchers (Wolfson, 1981;Homes & Brown, 1987), languages differ from each other in terms of speech acts and their linguistic realizations.These differences in speech act conventions have been one of the main causes of cross-cultural misunderstanding.Comparing American and other cultures in terms of semantic and syntactic structure of compliment, Wolfson (1981) observed that there are differences among languages regarding how to compliment and what is complimented.Similarly, Holmes and Brown's (1987) study indicated that misunderstandings in compliment exchanges may arise between people from two cultural and linguistic backgrounds.Their study which was based on two major compliment topics of appearance and ability indicated that learners need to understand not only topic differences related to compliments, who to compliment and when, they should also learn about the cultural values that these differences might convey.In other words, learners should be given chance to pay due attention both to pragmalinguistic and sociopragmatic elements about compliments.
Comparing the way American and Japanese give and receive compliments, Diakuhara (1986) also identified a number of differences between two languages with regard to praise attributes, frequency, functions and compliment responses.The most striking difference between them was related to compliment responses.While Japanese speakers tended to avoid self-praise in order to create distance by showing deference, Americans were more oriented towards establishing solidarity by accepting the compliments given to them.Likewise, Han (1992) found out that Korean female learners of English differ in their responding to compliment responses whether they spoke English or Korea.They tended to reject compliments in English interactions by saying "thanks" or "thank you".Other researchers such as Falasi (2007) and Nelson, Al Batal and El Bakary (2002) investigated the speech act compliments and their responses in English and Arabic languages and they also found some nuance differences in the way two languages formulate this speech act.
Using discourse completion test (DCT), Yousefvand (2010) investigated and categorized the range of strategies used in responding to compliments in Persian.30 undergraduate students majoring in English-Persian translation from Islamic Azad University, Khorasgan and state University of Isfahan participated in the study.Their responses to different scenarios presented in DCT resulted in 540 compliment responses.The results of her study indicated that although Persian speakers who participated in her study tended to respond to the compliments given them with "agreement", they showed their agreement with some sort of modesty which is the characteristic of Iranian culture.She also noted that there is some variation across gender in responding to compliments in a sense that males were more likely to reject a compliment by using a set of formulaic expressions.
Finally, it should be noted that the main insight gained from all the studies for pedagogy is that teachers and syllabus designers ought to make pragmatic features accessible to foreign or second language learners through pedagogical intervention (implicit or explicit instruction) or by incorporating this aspect of language into textbooks or other materials.The rational underlying this claim is Schmidt's noticing hypothesis that learners would turn input into intake only when they have conscious awareness of the target elements they want to acquire.

Participants
The people who participated in this study came from different parts of Iran including Shiraz, Tehran, Mashhad, Arak, Shahre Kord and Taybad.We tried to give compliments to people from different regions, occupations, levels of educations and age levels so that the participants could be as representative of the target population ( Iranian people) as possible.213 of the participants were male and the rest (40 percent) were female.The sites from which participants were chosen were as diverse as possible as follow: parks, academic settings, family gatherings, markets, dorms and virtual environments (online chatting).About half of them were below 30 and the rest were higher than 30.In order to add variety to the compliment responses, 12 other people helped the researchers to gather data.In other words, they were also asked to give compliments to different people.Of these 12 assistants, two were PhD students, four businessmen, four teachers and the rest housekeepers.All those who were about to give compliments knew in advance the topics of compliment e.g.ability or appearance so that a greater variety of compliment responses could be elicited.In other words, the researchers and their assistants went about giving compliments while knowing what topics to cover.

Procedure
This study is mainly based on naturally occurring conversations performed by researchers and their assistants as those who gave compliments and the various participants who provided responses to compliments.The conversations were either audio-tapped and then transcribed or were written down immediately after the conversations so that the conversations could be retained in a verbatim manner.In order to achieve a comprehensive entry of compliment responses in Persian, attempt was made to give compliments on a variety of topics including mental ability, possession, acts and nationality.Moreover, the range of syntactic formulas used for giving compliments was considered as well.We tried to include almost all the major syntactic formulas in Persian to account for the effect of grammar on the types of responses given to compliments.In order to obtain the full potential of compliments responses, we attempted to ensure a fair distribution of both syntactic formulas as well as topics of compliments based on Rose and Ng Kwai-fun (2001).
Since giving compliments to somebody requires some degrees of acquaintance, we had to use the same 12 assistants mentioned in the previous section to help us in this study.In this way, we could maintain the naturalness of the conversations without resorting to contrived situations in which participants were informed in advance that they are going to be given a compliment, and to prepare themselves to answer appropriately.Instead, every one of us could give compliments to people on the other side of the conversations who were among his or her friends, siblings, professors, colleagues or students.In case of researchers' assistants who could not or did not want to use tape-recorder, one of the researchers accompanied them to write down the conversations while occurring or immediately after it.

Results
As mentioned before, the corpus of this study is comprised of 756 naturally occurring compliment responses in Iranian context.By naturally occurring, it is meant those compliment responses obtained not through questionnaires (e.g.DCT) which do not directly measure the social or linguistic actions.In other words, because of the artificiality of the data obtained through discourse completion test, we tried to maintain the naturalness of the data by giving different types of compliments to various groups of people orally in different contexts.Herbert's (1986) taxonomy of compliment responses was the primary framework for coding them.Furthermore, two types of complement responses from Miles (1994), three types of compliment responses from Rose and Ng kwai-Fun (2001) and finally one type from current study were added to Herbert's taxonomy.Overall, the final classification of compliment responses ended up in 18 types.The first 12 types are adopted from Herbert's (1986), the next two (types 13 & 14) from Miles (1994), the next three ones ( types 15, 16 & 17) from Rose and Ng kwai-Fun ( 2001) and the last one (type 18) from current study.Each of them is described below and some of them are provided with an example.Although the study of compliment responses in Iranian context has been carried out before, this study makes use of data elicited through naturally occurring conversation instead of DCT.
1. APPRECIATION TOKEN.A verbal acceptance of a compliment, acceptance not being semantically fitted to the specifics of that compliment, 2. COMMENT ACCEPTANCE.The addressee accepts the complimentary force by means of a response semantically fitted to the compliment.
3. PRAISE UPGRADE.The addressee accepts the compliment and asserts that the compliment force is insufficient.
4. COMMENT HISTORY.The addressee, although agreeing with the complimentary force, does not accept the praise personally; rather, he or she impersonalizes the complimentary force by giving (maybe irrelevant) impersonal details. 5. REASSIGNMENT.The addressee agrees with the compliment, but the complimentary force is transferred to some third person or to the object complimented itself.6. RETURN.The praise is shifted to the addresser/complimenter 7. SCALE DOWN.The force of the compliment is minimized or scaled down by the addressee.8. QUESTION.The addressee might want an expansion or repetition of the original compliment or question the sincerity of the compliment.9. DISAGREEMENT.The addressee directly disagrees with addresser's assertion.10.QUALIFICATION.The addressee may choose not to accept the full complimentary force offered by qualifying that praise, usually by employing 11.NO ACKNOWLEDGEMENT.The addressee gives no indication of having heard the compliment; that is, he or she employs the conversational turn to do something other than responding to the compliment offered, e.g., shifts the topic.
12. REQUEST INTERPRETATION.The addressee interprets the compliment as a request rather than a simple compliment.18. AVOIDING EMBARRASSMENT.The addressee feels embarrassed to be complimented.(e.g.Don't put me to shame, I didn't do something extraordinary).
In order to examine the potentials of different data collection procedure, the results of this study are compared to those of Yousefvand (2010).Additionally, a comparison will be made with Herbert's study to find out about the potential differences across cultures.However, it should be noted that due to short of space, only the percentages of those compliment responses that we want to compare or contrast are presented without referring to the complete list of their compliment responses.For more information regarding their findings, readers can refer to Yousefvand (2010) and Herbert (1986).
As shown in table 1, agreement responses in our data amount to 67.6% in the compliment exchanges.47.6% of this belongs to the subcategory "acceptance" and the rest (20%) makes up the subcategory "non-acceptance".So, it is quite evident that the most frequent type of compliment response among Iranians is "agreement" and among this main category, the subcategory of "acceptance" roughly makes up half of all compliment responses.This strongly suggests that Iranians are more likely to accept rather than reject a compliment.Although the main category "agreement" was also the most frequent type of compliment response in Yousefvand ( 2010) study (43.49%), this value is much higher in our study, and it makes up two-third of the total responses.This finding is similar to that of Herbert's study on compliment responses among American and South African college students.Agreement responses in his study comprised 66% of all the compliment responses which is roughly equal to our study (67.6%)."Non-agreement", the second main type of compliment response makes up 24.2 % of the compliment responses in our data which is quite similar to Yousefvand's study in which disagreement comprised 24.82% of the total responses to compliments.Finally, the "other interpretation" which consists of "request" according to Herbert's [making offer according to Rose and Ng Kwai-Fun(2001) comprised 7.9% of our data.In Yousefvand's (1990) study category "other interpretations" which consists of requests as well as formulaic expressions, comprised 31.7% of the responses.In our view, "formulaic expressions" cannot be considered as a different type of compliment responses since they appear in almost all types of compliment responses, and they have nothing to do with the semantic content and illocutionary force of compliments.For example, the following compliment responses contain formulaic expressions: (in all the examples, the italics are Persians and the non-italics are their English equivalents).
A: What a beautiful cell phone!B: I'm sacrificed for you.You are welcome.

3) Agreement: A: In sha allah yeki az behtarin khalabanhaye Iran mishi! B: Merci, Shak nadaram.
A: God willing, you will make one of the best pilots of Iran.B: Thanks, I have no doubt.
A: Shirazi's girls are very beautiful!B: You are right, don't ever doubt it.
The main point is that formulaic expressions in contrast to Yousefvand's study do not make up a separate category or type in compliment responses, rather they only accompany other types of compliment responses and their roles according to Yousefvand (2010) is to show their modesty.
Another aspect of the analysis focuses on the differential responses given by males and females.The quantitative analysis of the participant's responses to compliment in terms of the three main types [agreement, non-agreement and other interpretation] revealed close similarity across genders.Table 2 shows that males and females' compliment responses consisted of 66.5% and 68.48% of agreement type respectively.Similarly, 24.9% of males and 23.3% of females rejected the compliments given to them (males 8.2% while females 7.6%).Thus, it seems that there are not significant differences between males and females in terms of compliment responses.In other words, the preference for acceptance of the compliment applies rather equally to both genders.The third way to analyze the data is to examine the frequency of different response types in terms of the topic of compliments.For this purpose, different topics of compliment were reduced to three main topics including physical compliments, mental compliments and compliments given on participants' nationality [ e.g. to say something nice about participant's hometown, for example, the Taybadi's are very hospitable].As table 3 displays, the overall tendency in all topics is toward acceptance but the interesting point is related to the frequency of the compliment responses given on nationality.While in compliments related to topics of physical and mental, 64% and 63% of participants [about twothird of them] respectively showed a preference towards acceptance, in case of nationality, about 90% of participants agreed to compliments.

Discussion.
Concerning the frequency of different response types in Persian, it can be argued that "agreement" comes at the top of compliment responses with appreciation token with the highest frequency (19.3%).The main characteristic of this type of compliment response is that at the same time that the complimentee accepts the illocutionary force of compliment, he or she tries to accompany this type of response with one or two forms of what Yousefvand (2010) calls formulaic expressions in order to avoid self-praise, or show modesty by analyzing the strategy of "shekaste nafsi" [humbling oneself] (Sharifian, 2005).Thus, it can be inferred from our data that the response type of appreciation token which makes up 19.3% of all responses is in line with Leech's (1983)  As far as the compliment responses of "appreciation token" is concerned, the results of our study is in line with those of Yousefvand's (2010) study in a sense that the use of formulaic expressions such as "chakeretam"[ I am your servant], "nazare lotfe shomast" [ its very kind of you] and " khejalatam nade"[ do not put me to shame] give support to cultural scheme of " shekaste nafsi" put forth by Sharifian (2005).However, it should be noted that appreciation token only makes up 19.3% of the total responses subsumed under the main category of agreement.Within this main category, other subcategories including comment acceptance (6.1%), praise upgrade (7.9%), agreement (10.8%), continual effort (1.4%) and nonverbal response (2.1%) amount to 28.3%.In other words, in these subcategories of "agreement", there is no sign of modesty, and it seems that Iranians, contrary to the expectation, accept the compliment comfortably.The following example might clarify this point: A: Hushe Riazie ballayi darid!(You have an amazing mathematical intelligence) B: Khoda'am!(I'm a god in mathematics).
A: Range kif va mantotoon ro kheili khoob set kardid! (The color of your bag goes very well with that of your dress) B: Daghighan.(Exactly) This claim can further be supported by considering the compliment response "praise upgrade".While Herbert and Yousefvand's data consists of 0.04% and 4.80% of praise upgrade response, our data is consisted of 7.9% of praise upgrade response type.This is in stark contrast with those studies that claim that Asian people tend to reject compliments, or lessen embarrassment and tensions between interlocutors (Chen, 1993& Loh, 1993).
It is worth discussing the fact that a new type of compliment response emerged out of our data which we call "avoiding embarrassment".Although this type of compliment makes up only a tiny fraction (3.3%) of our data, it lends support to Sharifiyan's (2005)  A mention is worth to be made of compliment response of "reassignment" in our data.As can be seen in table 2, reassignment only accounted for 2.2% of all compliment responses.Of this percentage, only a handful of compliment responses were assigned to God.
With regard to non-acknowledgement (silence) as a non-agreement compliment response, no respondent remained silent after receiving a compliment, however, this does not mean that the participant did not reject compliments; rather they resorted to verbal responses such as scale down and question.This might be interesting in a way that in Iranian culture, remaining silent as a sign of rejecting a compliment may not be considered a polite way of responding, although remaining silent along with a facial expression such as smiling as a way of accepting the compliment is not far from expectation (2.1% in our data).Nonetheless, this finding is in contrast to both Yousefvand's (2010) and Herbert's (1990) whose data consists of 4.08% and 5.1% of non-acknowledgement response type respectively.This discrepancy might be related to the nature of data obtained in these studies.Whereas in our study, participants got engaged in face to face interactions, they might have felt more obliged to utter something no matter what, the participants in the aforementioned studies only responded to imaginary situations depicted in DCT, and hence, they felt less obliged, if any, to respond.
Another interesting finding in our data is the remarkable amount of request response [or making offer] which amounts to 7.9% of the total compliment responses.To be more precise, making offer comprised 23% of the compliment responses when the topic of complimenting was physical [e.g.possession such as cars, clothes…].This finding runs counter to the studies of Herbert's (1990) and Yousefvand (2010) in which request [making offer] makes up only 2.9% and 0.92% of all compliment responses.This tendency among Iranians to make offer when they are given a compliment on their possessions can be explicated in light of the concept of "Ta'arof" in Iranian culture.According to Sahragard (2004), Ta'arof is a part of Iranian culture which shows Adab (politeness), Tavaazo (humility), Ehteram (respect), Rudarbaayesti ( being shy or ceremonious ), and finally MehmanNavaazi (hospitality).Looked at from this cultural script approach taken by Sahragard, we come to this conclusion that Iranians would enact one of the above realizations of the concept of Ta'arof when they are given a compliment on their possessions.The following examples better clarify this point: As mentioned before in the result section, no significant differences were found between men and women with regard to the frequency of compliment response types.Both males and females showed similar pattern of compliment responses in different topics of compliment.Of 18 types of compliment responses, men differed from women only in compliment response of agreement (males 7.7%, females 14%) and disagreement (males 10.7%, females 4.9%).The stronger preference of females for compliment response of agreement (9.8%) can be interpreted in two different ways; Women are either more likely to accept compliments, or its better to say they like to be complimented, or they want to avoid threatening the positive face of those people who compliment them.According to Yousefvand (2010), "the female participations seem to regard the denial response pattern as an inappropriate option".Thus, our study indicates that gender is not a defining variable in the compliment responses; that is, gender disparity across compliment responses is not significant.
With regard to the topic of compliments , all the compliment responses were classified into three main topics, namely, physical [ appearance, clothes, possessions], mental [ intelligence, behavior, character] and nationality [ e.g. to say something nice about somebody's hometown].Of the total number of compliments (756), 343 involved the physical topic, 189 involved mental topic and the rest (102) involved nationality.While the participants' responses to the first two topics were similar in terms of the two major response type of agreement and non-agreement; the third topic brought about quite different results.As displayed in table 3, agreement response accounted for 64% of physical topic, 63% of mental topic and 89% of nationality.To be more specific, only 11% of the participants who were given the compliments on their nationality refused to accept it.This indicates that Iranians are passionately sympathetic to patriotic causes.In other words, although the general response pattern among Iranians is more inclined towards agreement end of the continuum of compliment response, in case of nationality topic, this inclination is much more conspicuous.Furthermore, most of those responses related to nationality fell into the subcategories of "agreement" and "praise upgrade".The following example better clarify the point: A: Shoma birjandiha adamhaye khoongarm va mehrmannavazi hastid!B: Chizi ke ayan ast che hajat be bayan ast.
A: You Birjandis, are very friendly and hospitable!B: That, which is clear, does not warrant saying.

Conclusion
The above discussion suggests that, generally speaking, the Iranians' responses to compliments tend to lean towards agreement and appreciation.This in turn rejects the finding of those studies (Chen, 1993;Loh, 1993& Othman, 2011) which imply that the Asians tend to reject compliments.This however, might come true with south-east Asians not with south-west Asians, specifically, Iranian people.Compliment responses tend to correspond to those of American people studied by Herbert in terms of the main categories of agreement, non-agreement and other responses.Though, this does not mean that these two cultures do not differ from each other.Among the discrepancies between the two cultures are the following: In our data, the rate of "request response" [or making offer] was much higher than that of Herbert's study; that is, 7.9 % versus 2.9% respectively.As stated before, making offer in Iranian culture is one way through which people make "Ta'arof".According to Sahragard (2004), Ta'arof plays a key role in establishing good and smooth relationships in a number of ways.Additionally, the appreciation token which comes at the top of compliment response types in our study [19.3%] can be interpreted from a perspective which is particular to Iranian culture.The majority of these appreciation tokens consisted of two or three parts.In other words, while accepting the illocutionary force of the compliment, in one part of the response, they said something to show modesty which is a key component of Iranian politeness (Yousefvand, 2010).
Two other outstanding features in our data concern the response types of non-acknowledgement and avoiding embarrassment to the absence of the former and the presence of the latter in Iranian culture again testify to what can be referred to as Iranian specific politeness.Moreover, as the result of the study indicated, there exist some differences between data collected through DCT as used by Yousefvand with regard to compliment responses and the data collected naturally in our study.However, caution must be exercised regarding this conclusion since the differences found between the two studies are not much significant.
In sum, the findings of this study are in agreement with those of other studies concerning the fact that languages differ from one another in terms of speech acts and their linguistic realizations (Wolfson, 1982a &Holmes andBrown, 1987).This pragmatic variation among languages testifies to the fact that language is not an object devoid of social and cultural elements, but it is a process which is inseparable from the rest of human social life (Haliday, 2003).The implication for the L2 teaching is crystal clear; speech acts and other pragmatic features of L2 should be high on the agenda.Both teachers and material developers are strongly recommended to pay more attention to this aspect of language.
. The addressee aligns himself with the complimenter (e.g.I like it too) 14.SELF-PRAISE AVOIDANCE.The addressee avoids being flattered (anyone can do this) 15.CONTINUED EFFORT.The addressee ascribes the point of compliment to his effort not something like a given talent.(e.g.I will keep doing it) 16.NON-VERBAL RESPONSE (silence to show agreement) 17.ANSWER QUESTION.The addressee interprets the compliment as a question to which he provides an answer.(e.g.I bought it last week).

Table 1 .
Frequency of Compliment Response Types among Persian Speaking participants

Table 2 .
Frequency of Compliment Response Types across gender

Table 3 .
Frequency of Compliment Response Types in terms of topic of compliment cultural schema of shekaste-nafsi [modesty].Let's take a look at the following examples: A: che sofreye por zargho barghi darid!B: khejalatam nade toro khoda, engar ke gav koshtam baratun.A: What a generous table you have set!B: Do not embarrass me please, as if I have slaughtered a cow for you.