Complimenting is considered a complex sociolinguistic skill (Holmes, 1988). Compliment
speech act is argued to be “worthy of study because it is ubiquitous, valued, and problematic”
(Knapp, Hopper, & Bell, 1984, p. 12). Holmes defined the compliment imperfectly—Holmes
did not take into considerations the ‘negative aspects’ (such as insults) of complimenting in
her definition and also she mentioned that compliments are ‘positively valued’ which is not
generalizable to some cultures such as Polish— as “a speech act which explicitly or
implicitly attributes credit to someone other than the speaker, usually the person addressed,
for some ‘good’ … which is positively valued by the speaker and the hearer” ignoring the
negative aspects of compliments (Holmes, 1988, p. 446). Likewise, Nkwain (2011) defined
compliments, much more inclusively, as “a positive politeness strategy that expresses
goodwill and solidarity between interlocutors, although compliments tend to serve other
functions, depending on the interpretation they are coded” (p. 61). The perplexing issues
surrounding the compliment speech act are not as easy and clear-cut as the definitions put
forward by Holmes and Nkwain. Compliments function differently vis-à-vis individuals and
cultures throughout the world. Many Arab people, for instance, believe that compliments
would invoke the evil eye (Nelson, El Bakary, & Al Batal, 1993). For New Zealanders, it is
considered inappropriate and rude to compliment a man on his wife since the illocutionary
force of the compliment view the wife as a possession (Holmes & Brown, 1987). In
American culture, compliments act as a social lubricant, albeit formulaic, and are performed
to create rapport among individuals (Manes & Wolfson, 1981). For Polish speakers, as
Jaworski (1995) argued, compliments are considered to be insincere and are interpreted as
“purely social act[s]” (p. 70). In other words, Polish speakers reject the complimentary force
of sentences. In Poland, therefore, compliments are interpreted as both a positive politeness
device as well as a threat to the addressee’s negative face since they may imply a desire for
the addressee’s possession or trait.
Previous works on compliments illustrates that context plays an essential role in
figuring out the true interpretation of compliments. The influence of compliments on social
success is compared with the role of oxygen in breathing (Knapp et al., 1984). Compliments
are largely subconscious (Wolfson, 1981) and it is necessary to collect and analyze
compliment patterns across cultures to avoid cross-cultural misunderstanding. Compliment
can also be direct or indirect and requires interpretation beyond linguistic norms.
The purpose of this study is to compare complimenting patterns between native English
speakers and Iranian EFL learners with regard to topics, functions, gender differences and
adjective types used in utterances. To this end, the following detailed research questions are
(1)What are the differences between compliment functions produced by native English
speakers and Iranian EFL learners with regard to gender?
(2) What are the differences between compliment functions produced by native English
speakers and Iranian EFL learners with regard to topics?
(3) What are the differences between native English speakers and Iranian EFL learners with
regard to the common positive adjectives used in compliments?
A general question of this paper would be to address whether the language system of
Iranian EFL learners approximates more to L1 or L2 norms. TESOL programs are now
widely practiced in Iran. In this multicultural country, members of speech communities
interact with each other using various sociocultural and linguistics norms. Therefore, it is
more likely that Iranian EFL learners would follow specific conventions of their L1 to
produce sentences in L2. On the other hand, there might be some divergence from the first-language norms. The learner language may be more inclined towards the L2 culture being
influenced by media or textbooks that are produced by English speaking countries. In this
way, Iranian EFL learners’ utterances may be more similar to L2 norms, and speech acts used
by learners are among the first manifestation of this influence. Therefore, a comparison
between the complimenting patterns performed by Iranian EFL learners and native English
speakers would be fruitful in determining the learners’ preference for convergence or
2. Literature Review
As mentioned by Boyle (2000, p. 26), “of all the speech acts studied by researchers in
sociolinguistics, pragmatics, and discourse analysis, few can have received more widespread
attention in the past 20 years than compliments”. The vast literature reveals that compliments
are studied according to the function(s) they play in interactions (Czopp, 2008; Maíz-Arévalo
& García-Gómez, 2013; Mustapha, 2012; Wolfson & Manes, 1980), gender differences
(Herbert, 1990; Holmes, 1988; Rees-Miller, 2011; Wolfson, 1984), cross-cultural comparison
(Chen & Rau, 2011; Maíz-Arévalo, 2012; Sharifian, 2005; Wolfson, 1981), and compliments
at the workplace (Hudak, Gill, Aguinaldo, Clark, & Frankel, 2010).
2.1. Functions of the Compliments
Among the current studies done on compliment speech act, most of them contributed to
determine various functions of compliments. The study conducted by Wolfson and Manes
(1980) is regarded as the pioneer one, considering the investigation of compliment functions.
Although Wolfson and Manes (1980) believed that compliments can have various functions
within a conversation such as, inter alia, a greeting or offering a topic for conversation,
American compliments have the primary role of reinforcing solidarity among the
interlocutors (Wolfson & Manes, 1980). Knapp et al. (1984) found out that compliments are
categorized into four main dimensions, namely, direct/indirect, specific/general,
comparison/no comparison, and normal/amplified. It was also revealed that 89 percent of 396
American compliments are categorized under the direct form. In an interesting study on
Polish compliments, Jaworski (1995) proposed that Polish speakers do not maintain solidarity
by means of compliments. Compliments are considered as a purely social act and are
interpreted as insincere praise. Polish speakers may even regard compliments as an act of
cheating. Polish compliments have multiple functions such as reinforcing desired behaviour,
congratulating, information seeking, and teasing.
Some researchers, however, were not much satisfied with the quality of studies done on
compliment functions. Boyle (2000) critically mentioned that “a more balanced picture of
complimenting is required and that the neglect of the study of implicit compliments should
not continue” (p. 26).
After these groundbreaking works, many other studies were conducted to find out the
functions of compliments in different cultures. Czopp (2008) mentioned that compliments
might not be welcomed by some speech communities, such as African Americans, because
compliments include negative stereotypes. Grossi (2009) investigated various functions of
compliments in Australian English. Similar studies are done on Cameroon Pidgin English
(Nkwain, 2011), Nigerian English (Mustapha, 2012), Iranian EFL students (Sadeghi & Zarei,
2013), and Japanese speakers (Kondo, 2014). In all of these studies, an attempt was made to
find the function(s) of compliments in different situations.
2.2. Gender Differences in Compliments
A large portion of research on complimenting is devoted to the role of gender in giving and
receiving compliments. The first attempt to gather empirical data on gender differences in
compliment speech act was made by Wolfson (1984). She reported that women use
adjectives such as adorable, charming, sweet, lovely, and divine most often in their
compliments. More interestingly, it was understood that women receive the great majority of
compliments both by male and female speakers. The most controversial issue that Wolfson
mentioned and contradicted sharply with Lakoff’s (1973) argument, is that “the way a
woman is spoken to is, no matter what her status, a subtle and powerful way of perpetuating
her subordinate role in society” (p. 243). Holmes (1984) endorsed that compliments have
various functions in men and women conversations and further mentioned that females are
more apt to give and receive compliments (about 51%). She associated this feature to the
“women’s positive attitude to compliments” (p. 451). For men, as Holmes argued,
compliments do not function to maintain solidarity, as was the case for women. However, in
2002, Mojica’s findings provided counter evidence to Holmes’s findings. Mojica found out
that Filipino males’ compliments have the primary role of establishing solidarity with the
females but “females want to assert their power in language” (p. 123). Another contradiction
in the compliment literature was addressed by Rees-Miller (2011). Unlike what was proposed
by Wolfson and Holmes, Rees-Miller (2011) suggested that in goal-oriented settings, men
give and receive more compliments than women. Nevertheless, women complimented on
appearances more often. Women dominated men in giving and receiving compliments in
2.3. Common Adjectives and Compliments
After extensive literature review, it was understood that there is paucity of research regarding
the investigation of adjectives among the appraisals. The most well-known are the studies
done by Wolfson (1981, 1984). The amount of seventy-two adjectives was identified in
compliments and five types of adjectives were recognized among the appraisals, viz., nice,
good, beautiful, pretty, and great. Nice (22.9%) and good (19.6%) were observed most
frequently in the data.
Wolfson (1984) mentioned that nice is the most commonly used adjective among
American English speakers (23 per cent). In Wolfson’s (1984) study, good was observed
more often after nice (20 per cent). The sequence of common five adjectives in compliments
according to frequency as revealed by Wolfson is nice, good, beautiful, pretty, and great.
2.4. Current Study
Literature review showed that there is paucity of research with regard to the analysis of
compliment utterances performed by native English speakers and Iranian EFL learners
through a written Discourse Completion Test. Most of the studies in the field have employed
different participants or instruments. Sharifian’s (2005) research is one of the best-known
studies which investigated the compliment responses of Persian speakers of Australia and
native Australian speakers. In recent years, compliment speech act has gained the attention of
the researchers. Behnam and Amizadeh (2011) analyzed compliments and compliment
responses in TV interviews manifested by Persian and English interlocutors. Likewise,
Karimnia and Afghari (2011) used TV interviews to study compliment responses of native
Persian speakers and Native American English speakers. In addition, more recently, Sadeghi
and Zarei (2013) studied compliments produced in Persian and English by a group of Iranian
EFL learners. The current paper is unlike earlier studies done on Persian compliments with
regard to the variables of this study. Precisely, this study focuses mainly on functions of the
compliments, and topics and gender as two moderating variables.
In pragmatic studies, data can be elicited empirically or can be gathered naturally through
conversation analysis. Each method has advantages and disadvantages (Bardovi-Harlig &
Hartford, 2005). In this contrastive study, a descriptive research design was used. The type of
research questions and data analysis used in this study pairs with some of the underlying
elements of descriptive design.
The completed forms were received from two groups, namely, 60 advanced Iranian EFL
learners who were speaking Persian as their first language and 60 native English speakers. As
argued by Kasper and Dahl (1991), for studies which employ a Discourse Completion Test as
the main instrument, it is considered appropriate to use a sample size of at least 30
participants. Similar works have included less than 60 participants in their study (Chen &
Rau, 2011; Lin, Woodfield, & Ren, 2012; Sadeghi & Zarei, 2013; Sharifian, 2005). In this
study, the participants were equally distributed with regard to gender. Other variables such as
educational level, field of study, and age were also recorded for further analysis.
Iranian participants were mainly postgraduate students and a small number of them
were undergraduates from different universities in Iran, aged between 21 and 50 years. All of
the Iranian participants were studying English Language Teaching, Translation Studies, or
English Literature as their main field of study. It was preferred to use postgraduate students
to ensure that the situations provided in surveys are understood completely. Undergraduate
students who were in the final year of their program were allowed to contribute.
Native English speakers were mainly university students (26 non-degree participants or
high school level, 33 BA/BS level, 6 MA, 1 PhD). Ten participants out of 66 cases were aged
under 20, thirty two participants between 21 and 30, sixteen participants between 31 and 40,
and eight cases over 40. They were volunteer participants from different English-speaking
countries. One reason to choose participants from different English-speaking countries was
lack of native speakers’ availability. It is true that there may be cultural differences among
individuals but with regard to compliment speech act, there might be some universals among
native English speakers.
Moreover, although Iranian EFL learners and native English speakers do not share the
same characteristics, comparing these two groups would help the researchers to find out
whether advanced Iranian EFL learners’ linguistic patterns are similar to or different from
native speakers’ norms. In pragmatic studies, appropriateness is the key factor and native
speakers’ compliments are regarded as a norm to be compared with Iranian EFL learners’
A written Discourse Completion Test was used to elicit data from Iranian EFL learners and
native English speakers. Appendix A has manifested the DCT employed in this study. The
application of DCTs in pragmatic and discourse studies has not been without controversies.
They have been criticized for their non-interactive nature (Bardovi-Harlig & Hartford, 2005),
excluding non-verbal features of communication (Kasper, 2008), and inability to record
subtle features of face-to-face conversations (Kasper, 2008).
However, despite the criticisms levelled against DCTs, there have been arguments for
the administrative advantages of them. Golato (2003) discussed that the DCT allows “the
researcher to control for certain variables … and to quickly gather large amounts of data
without any need for transcription, thus making it easy to statistically compare responses
from native and non-native speakers” (p. 92). The decision to gather data through the DCT
fits appropriately with the purpose of this study. It allows the researchers to have control over
variables and is time effective.
The DCT was designed in English and two native English speakers and a specialist in the
field of pragmatic research proofread the content both grammatically and semantically.
Content validity of the tests was checked by the use of subject matter experts. The DCT was
then modified based on the received feedbacks. There were eight situations provided in the
DCT and the participants were required to write down suitable compliments for each
situation. Scenarios were developed in a way to address four different common topics of
compliment, i.e., appearance, possession, skill/ability, and personality. That is to say, for each
topic, two scenarios were provided. In addition, respondents were asked to provide
information about their age, gender, place of birth, education level, field of study, native
language, and email address. Power relations and the distance between individuals were
excluded from the study since they were not included in the research questions and were
beyond the scope of this work. Respondents were also allowed to provide comments about
the DCT. A synopsis of comments is provided in Appendix B.
Participants were informed about the survey and purpose of the study mainly through
two venues: 1) InterPals that is a website for finding friends throughout the world, and 2)
Facebook. The questionnaire was posted online on Google Docs and the completed
questionnaires were automatically sent back to our mailbox.
Participants were required to provide written responses for each of the eight situations.
The situations used in the survey were neutral regarding gender. In other words, it was not
mentioned whether the participants should compliment a man or a woman. They were simply
asked to mention how they would provide a compliment in different situations. It is true that
some respondents may not have experienced the situations given in the survey but they all
may know the norms through which they should provide the most contextually appropriate
response. Scenarios required individuals to give compliments to people of equal status. Table
1 reveals a synopsis of the scenarios used in the DCT as well as the related compliment topic.
3.4. Coding System
The following coding system is adapted from Lin, Woodfield, and Ren (2012, pp. 1491-1492) and some minor changes were made with regard to the addition of micro strategies to
make the classification more inclusive. Four micro strategies were added to the original work.
These micro-strategies are no acknowledgment, appreciation, reciprocation, and positive
non-verbal comment which were absent in the original coding system. Explicit compliments
are characterized as being generally direct and unambiguous positive statements that contain
at least one positive semantic carrier, for instance, pretty, great, nice, good, etc. Implicit
compliments are those remarks that do not possess a positive lexical item as manifested by
linguistic form, thus need more inferences from the interlocutor to reconstruct the implied
meaning. In addition, in opt-outs, individuals decide not to participate in providing a
The data were analyzed and the compliment sentences that were gathered through the
DCT were classified under the three generic categories of Explicit, Implicit, and Opt-out.
Another rater who was familiar with pragmatic studies and was informed about research
questions and methodology coded the same data. The Cronbach Alpha reliability statistic
showed the inter-rater reliability of .77, though not perfect yet acceptable. The coding system
that is used in this study is presented in Table 2.
Both descriptive and inferential statistics were used to analyze the compliment
sentences. At the first phase, the frequency of the macro- and micro-functions as well as the
percentages was identified. Each compliment sentence was assigned a function (Explicit,
Implicit, or Opt-out) and the number of different functions in each group was considered as
the total frequency of data in that group. For instance, the sentence, ‘That haircut suits you so
much. I love it’, has two complimentary sentences that should be assigned a function. The
first sentence is ‘Explanation’ (That haircut suits you so much) and the second sentence is
‘Admiration’ (I love it). Therefore, there are two implicit compliments. The rest of the data
were coded as such. In the second phase, Pearson Chi-square test was utilized to find out the
significance of differences with regard to the two groups of participants.
4. Results and Discussion
In this section, the data are analysed according to gender distribution, topics, and the common
adjectives observed in compliment sentences. The frequency distribution of data as well as the
percentages is provided. To find out significance of differences, Chi-square test is utilized
4.1. Compliment Functions Based on Gender
The classification of compliment functions into three main generic categories (Explicit, Implicit,
Opt-out) based on gender differences is provided in Table 3. Findings revealed that explicit
function is the most common type of compliments by both Iranian learners and English speakers.
The most frequent macro function among Iranian EFL learners is related to explicit compliments
(f= 333). Iranian female EFL learners, however, provided more compliments than men with regard to
total frequency (335 cases vs. 323 cases). Male EFL learners are inclined to use explicit
compliments more frequently (53%) than female learners (49%). For female learners, the most
common macro function is related to explicit compliments (f= 163) and the least used function is
opt-out (f= 25) as was the case with male learners’ use of opt-outs. The analysis of data by Chi-square showed that gender and macro functions are independent of each other among Iranian EFL
learners’ compliments (χ2= 3.256, sig.= .196, p<.05). Therefore, it can be interpreted that gender
does not have an important role in the distribution of compliments among Iranian EFL learners.
Gender has not been proved to be a source of pragmatic variation among Iranian male and female
Female learners provided more ‘positive non-verbal comments’ in their compliments
than male learners. It was expected to observe the reverse trend since women were believed
to have a more conservative role in the Iranian culture. Thus, the results of Holmes’s (1988)
study are reinforced regarding the fact that women have positive attitude to compliments.
Iranian female learners also provided more opt-outs and fewer acknowledgements in giving
compliments in comparison to Iranian male learners.
Like Iranian learners, the most frequent compliment function produced by native
English speakers is related to explicit category. English speaking men used explicit
compliments (51%) more than female English speakers (42%). Although explicit
compliments were dominantly used by male English speakers (51%), female English
speakers used implicit compliments more commonly (52%).
The results of Chi-square (χ2= 11.469, sig.= .003, p<.05) revealed that gender has a
significant effect on the differences observed in macro functions among native English
speakers’ compliments. Gender has a significant role in the distribution of compliments
among English speakers. Therefore, the second hypothesis of this study is rejected. There is a
significant difference between compliment functions produced by male and female native
To answer the first question of this study, as summarized in Table 3 and Figure 1, the
dominant macro function for both Iranian EFL learners and native English speakers is
explicit compliments, except for female native speakers. That is to say, just female native
speakers were more inclined to use compliments implicitly but other groups employed an
explicit strategy dominantly. Overall, female native English speakers also uttered more
compliments than other participants (f= 338). In this case, the results are in line with the
argument put forward by Wolfson (1984) and Holmes (1988). They proposed that English
speaking women give and receive the compliments more than men. Moreover, Holmes
mentioned that “women use compliments to each other significantly more often than they do
to men or men do to each other” (p. 462).
To find out the significant of differences between Iranian EFL learners and English
native speakers with regard to functions and origin (being an Iranian EFL learner or an
English speaker), the Chi-square test revealed a value of .185 which is non-significant (χ2=
3.372, p<.05). Therefore, being Iranian or English (in the case of our data), does not have a
pivotal role in the variances observed between these two groups with regard to compliment
functions. In other words, both groups had similar viewpoints about using different functions
In providing compliments, both native speakers and Iranian EFL learners provided
some comments which described their cultural norms and values and the way these
conventions would affect the complimenting patterns. In one situation, for instance, a
participant from New Zealand mentioned the following comment to compliment a driver for
his skillful driving:
(1) I would not compliment. I can’t really answer this one because I feel a cultural obstruction with
mention such things to servicemen. I’m not too sure why.
An Iranian female learner provided the following comment for complimenting a taxi
(2) I would give no compliment. I don’t want to start a conversation with a taxi driver, especially
admiration. I had bad experiences.
In complimenting a fellow in the bus, another Iranian female learner mentioned:
(3) No answer. I think there is no need to admire a person that I don’t know.
In Iranian culture, women would rarely compliment a stranger because of cultural
stereotypes; it may be regarded as ‘inappropriate desires’ in some circumstances. Thus, many
women avoid sharing a compliment with a stranger (a taxi driver can be a stranger). Eight out
of 15 cases in which Iranian female learners avoided to provide a compliment, were related to
complimenting a taxi driver.
4.2. Compliment Functions Based on Topic
Compliment functions have been analysed with regard to the four main topics of
compliments (Appearance, Possession, Performance/Skill, Personality). Table 4 shows the
distribution of functions based on topics performed by Iranian EFL learners.
As shown in Table 4, Iranian EFL learners favored explicit compliments for every topic
except for personality in which the implicit compliments (64%) overweigh other functions.
Compliments about appearance (66%) appeared most frequently among Iranian EFL learners.
Table 4 represents the distribution of compliments based on topics provided by native
As related to native English speakers, the same pattern was observed as Iranian EFL
learners though with different frequencies. According to Table 4, for compliments on
appearance, possession, and performance, the most used function is explicit. However, for
personality, it was implicit function (75%) which was used most frequently by native English
speakers. For possession and personality, participants provided more compliments than other
topics (f= 172).
To answer the second question of this study, for appearance, both Iranian EFL learners
and native English speakers provided explicit compliments more than other functions
(although Iranian EFL learners’ compliments outweigh English speakers ones). For
possession, native English speakers used more explicit compliments than Iranian EFL
learners. Regarding performance, Iranian learners produced more explicit compliments than
native English speakers. Finally, for personality, explicit function was the most dominant
type of compliment. Iranian learners produced more explicit compliment than English
speakers. Figure 2 shows the differences between compliment functions produced by native
English speakers and Iranian EFL learners with regard to topics.
As manifested by Figure 2, direct compliments such as the following were observed
more frequently in situations related to appearance:
4 (a) You look great.
(b) You look nice today.
Likewise, for possession, Iranian EFL learners and native English speakers used mostly
explicit compliments but the latter group showed a higher tendency to use explicit
compliments. Compliments such as the following are among the most common ones:
5 (a) What a nice phone you have got.
(b) That’s an awesome jacket.
For performance, most of the English speakers and Iranian EFL learners also provide
explicit compliments but for the former group the observed frequency was higher. In this
situation, the following compliments are observed as the stereotype:
6 (a) You have got a pretty handwriting.
(b) You are really good driver.
For personality, however, both native English speakers and Iranian EFL learners
provided the most compliments implicitly although native speakers used more implicit
compliments such as the following:
7 (a) Thanks for your help (Admiration).
(b) If you have a gust, let me know to help you (Reciprocation).
(c) It’s crazy that we can get on so well after only just meeting (Explanation).
(d) Wanna grab dinner on me (Joke).
Opt-out compliments, in which there is no complimentary force, were mostly observed
in scenarios about possession. In this situation, Iranian EFL learners used opt-outs in 24 cases
while native English speakers employed this macro function in 30 cases. Therefore, it is rare
that one could see compliments such as the following among Iranian EFL learners or native
8 (a) Is it a new brand?
(b) How did you end up being able to write like that?
(c) New style!
4.3. Common Adjectives in Compliments
Compliment sentences were also investigated to find out the common positive adjectives. The
type of adjectives and the frequency of each were analyzed. In Table 6, the adjectives are
categorized according to gender of the participants. The superscript figures manifest the
sequence of the most three frequent adjectives. It should be mentioned that the number of
positive adjectives used in the corpus was virtually unlimited but Table 6 shows the most
common ones observed in the data. Other adjectives such as sweet, elegant, graceful, fresh,
interesting, chic, cute, modern, trendy, fantabulous, skilled, dexterous, smashing,
approachable, fabulous, attractive, wondrous, easy-going, social, super, expert, smooth, clear,
talented, splendid, pleasing, girly, magnificent, generous, fine, dandy, radiant, and helpful
were used only once or twice and were not included in the table.
To address the third question of this study, for Iranian EFL learners, the most common
adjective which was observed in females’ compliments was nice (f= 38). Likewise, for male
learners, nice was the most observed adjective (f= 41). English speakers manifested the same
pattern too. For both male and female native English speakers, nice was the adjective with the
most observed frequency.
According to Table 6, Iranian EFL learners employ more varied positive adjectives in
their compliments than native speakers. This knowledge is useful to recognize how
compliments in English speaking countries are regarded as formulaic. Most English speakers
used a common set of adjectives in their compliments and nice, good, great, and cool, among
others, are mostly observed in their compliments. Iranian EFL learners, however, used nice,
good, great, beautiful, and kind more than other adjectives.
With respect to positive adjectives, Figure 3 demonstrates that nice is the most used
adjective to show the positive semantic load. The results endorse Wolfson (1984) who
mentioned that nice is the most commonly used adjective in American English (23 per cent).
Regarding the most five common adjectives, as shown in Figure 3, among Iranian EFL
learners and native English speakers, nice was used with the most observed frequency.
Iranian EFL learners used beautiful with a frequency of 53 but native speakers used good in
their compliments in 34 cases. Great is ranked the third in both groups of participants. Iranian
EFL learners used good with a frequency of 27 and native speakers used cool with a
frequency of 30. Finally, the least observed adjective among Iranian EFL learners was cool
while native English speakers used beautiful in their compliments in 12 cases
In current study, great outweighed other common adjectives after nice. However, in
Wolfson’s (1984) study, good was used more often after nice (20 per cent). The sequence of
common five adjectives according to frequency as revealed by Wolfson is nice, good,
beautiful, pretty, and great. However, the results of this study revealed that participants used
cool in their sentences very often which was not included in the list of common adjectives
proposed by Wolfson. Pretty was observed rarely in the adjectives of current study and it was
not used commonly by Iranian learners or native speakers. Cool, however, was largely
utilized by English speakers rather than Iranian EFL learners. In addition, cool was observed
mostly in compliments about possession.
Despite the availability of a wide range of vocabulary to show positive evaluation, in
majority of the compliments used in this study, a restricted set of adjectives and other
semantic formulas was used. Regularity was observed only in some of the adjectives and
verbs. Therefore, adjectival compliments such as the following were observed numerously in
9 (a) That’s a nice gadget you have got.
(b) You look beautiful today.
Although many of the compliments in this study were of adjectival type, some of the
compliments were comprised of a verb to manifest the positive semantic load. This type of
compliments was rare with comparison to adjectival type. The most common verbs used in
the compliments were like and love:
10 (a) I like your outfit.
(b) I love what you’re wearing today.
Adverbs, likewise, were applied to manifest the positive semantic load in compliment
sentences. For instance, well was observed sporadically in the compliments:
11 (a) Wow! You drive really well!
(b) You are a well-educated man.
Intensifiers are semantic elements which are used typically in compliments. The most
common intensifiers are, inter alia, ‘really’, ‘very’, ‘such’, and ‘so’:
12 (a) You are so sympathetic.
(b) I really like your new hair-cut.
Finally, deictic expressions are another common feature of compliment sentences. In
effect, ‘demonstratives’ and ‘second person pronouns’ are among the most observed deixis in
13 (a) This new hairstyle looks so good on you.
(b) That is some nice handwriting you have there.
The above analysis showed how Iranian EFL learners and native English speakers
produced compliments with regard to positive adjectives. Gender differences were also
investigated and the Chi-square test showed that gender has a significant role in the native
English compliments but not among Iranian EFL learners. Moreover, there is a significant
difference between Iranian EFL learners and native English speakers in producing
compliments with regard to gender. Origin, being an Iranian or an English speaker, also plays
an important role in the distribution of compliment utterances. Compliment functions,
however, do not have a significant effect on the compliment variation.
The result of this study demonstrated that there was no significant difference between the
way Iranian EFL learners and native English speakers produced compliments with regard to
functions. In other words, there were more similarities rather than differences between these
two groups of participants as was observed in complimenting functions. It was revealed that
both male participants in the two groups complimented explicitly. Iranian female learners
were also more inclined to compliment directly but female native speakers used compliments
more implicitly. Generally speaking, Iranian EFL learners provided more explicit
compliments (f= 333) than native speakers (f= 305). Native speakers were more inclined to
use compliments implicitly (f= 299) than Iranian EFL learners (f= 286). The results endorse
Brown and Levinson’s (1987, p. 248) argument that “indirect speech acts are highly
conventionalized in English means that in most circumstances using an indirect speech act
implicates that S is trying to respect H’s negative face.” Iranian EFL learners used less opt-outs than native English speakers. One justification for the dominance of explicit
compliments among Iranian EFL learners with comparison to native speakers is that Persian
culture highly respects the ‘positive face’ of people (positive face is the desire to be
appreciated and liked). Any attempt is made by Iranian EFL learners to show that the
compliment is understood and welcomed perfectly to avoid positive impoliteness (Culpeper,
1996), as positive impoliteness would breach the roles of an individual’s positive face. One
way to ensure the recognition of individuals’ positive face, as is the case with giving
compliments, is to resort to direct and unambiguous compliments. Similar findings are
provided by Behnam and Amizadeh (2011). They proposed that compliments are made and
responded to more often in Persian than in English. They further assigned this pervasive use
of compliments to the role of taarof in Persian culture.
Regarding topics of compliments, a similar pattern was observed between native
speakers and Iranian learners. Both of the groups, performed explicitly in all of the topics
except personality. Iranian EFL learners and native speakers used implicit compliments in
topics related to personality. Moreover, as related to adjectives, although both groups used a
variety of adjectives, nice was observed with the highest frequency in participants’
compliments. As a difference, Iranian learners used beautiful and great as the second and
third most frequent adjective respectively but for English speakers, good and great were the
second and the third adjective with the most observed frequency in compliment sentences
(refer to Figure 3). For native speakers, the results are consistent with Wolfson and Manes
(1980) which suggested that good, nice, and great are most frequently used in American
Therefore, to clarify the general question of the study, results suggested that the
‘approximative system’ of advanced Iranian EFL learners is more inclined towards the L2
norms rather than being interfered by first language structure. One possibility may be the
influence of media on learners. Thus, it may be possible that Iranian learners have picked up
target language norms (in this case English) and integrated them to their sentences.
This paper suggests that advanced Iranian EFL learners, rather than generating creative
appraisals, resort to a set of precoded formulas to provide compliments. Findings reject
Wolfson (1981) argument that Iranian speakers would include ritualized expressions and
proverbs in their compliments. Wolfson’s findings may hold true for native Persian speakers
but for the Iranian EFL learners it did not turn up to be true. In other words, Iranian EFL
learners are influenced by the appropriateness phenomenon. They try to include themselves
in the inner circle of native English speakers by adhering to the pragmatic norms of the target
language. One way to reach this goal is to provide formulaic expressions in different
situations. The results of this study also reinforce Knapp et al. (1984) ideas that about 89
percent of American English compliments are direct (or explicit). Briefly, complimenting in
Iranian EFL learners’ discourse is influenced by two factors: (a) to maintain a positive face
and (b) to respect a culturally valued norm of being polite. Complimenting in native English
speakers is not affected by emotion; the following comments, which are provided by some
native speakers, endorse this argument:
14 (a) I know that from a perspective of stereotypes, in general, people from the UK are ‘less
friendly’ to strangers, despite being a more collective society than the USA (Male from the
(b) We don’t tend to say things like ‘you were so kind and sympathetic’. Not sure why. I guess
many Americans don’t say ‘emotional’ things like that to most people. We might say ‘That
was so nice of you’ (Female from the United States).
(c) I can say with confidence, my circle of friends in New Zealand use compliments sparingly
among one another. With the males around me - I suppose similar things are true across
other cultures - they show compliments, less with words, than with gestures of brotherhood
(Male from New Zealand).
In this study, an attempt was made to include every possible detail that was related to
compliment speech act. It is important to say, however, that the participants of this study
(regarding Iranian EFL learners) were university students. The ordinary people may have
different complimenting patterns. Thus, care must be taken to generalize the findings to
overall population of Iran.
There are some areas of compliment speech at which have not been covered in this
study and can be regarded as the possible limitations. Compliment response patterns of
Iranian EFL learners and native English speakers have not been analysed in this study and it
would be a good idea should to investigate whether there is any particular pattern observed in
their responses. Moreover, distance and social status were excluded from this study.
Likewise, use of other data collection methods rather than a DCT, would provide important
results regarding the nature of compliments. Therefore, more studies should be done to fill
the possible gaps and limitations that still exist in the vast literature of compliment speech act.
The implications of the study are related to four areas. Firstly, researchers can benefit
from the results presented in this article to compare and contrast compliment speech acts in
other cultures to find out whether there is a universal pattern applicable to all cultures and
languages. The second implication is related to material developers. It is possible to prepare
educational materials to improve learners’ pragmatic competence. Studies like this can
provide authentic materials regarding the proper use of sentences in context. Teachers can
also benefit from the results of this study. Teachers relying on the results of this study, could
opt for providing authentic input and real language use for the learners. Most of the advanced
learners are grammatically competent and know many vocabularies but when it comes to
pragmatic competence, most of them will be challenged. Finally, the results of this study can
be useful for the learners who need to know how they can perform specific speech acts such
as compliments. Results suggested that people with different cultures perform particular
speech acts in various ways. Hence, being aware of these differences can help learners to
communicate more properly.
The authors would like to extend their gratitude to Dr. Abolaji Mustapha (Lagos State
University), Dr. Parisa Niloofar (Bojnord University), and the two anonymous reviewers for
providing invaluable feedback on the earlier drafts of this paper.