Overall, 420 strategies were produced across the four situations; 56 of these strategies were found in situations of Farewell and Advisor , while 44 were observed in situations of Birthday and Notes. Differences were observed with respect to strategy selection in the use of direct and indirect refusals and adjuncts to refusals between Persian native speaker and EFL learners. The findings suggest similarity in terms of types of refusal strategies, but differences in terms of frequency and the content of strategies used in Persian language of EFL learners and that of native speakers. As shown in Figure 1, a higher percentage for indirectness was found in almost the four situations for Persian native speakers and EFL learners. The direct refusal percentage, however, found preference more among EFL learners. Adjuncts had an increase in amount in almost every situation among EFL learners. Overall, indirect strategy got the most frequency among the two groups in every situation. This supports the findings of previous research at least in that refusals operate by universal rules in terms of preference of indirect over direct strategies in communications between the people who have ongoing relationship (see Kasper & Rose, 2003; Blum-Kulka et al., 1989).
Direct refusal or "Bald-on-record" refusals (in the terminology of Brown and Levinson, 1987) like 'no' 'I can't' or 'I won't' were observed very occasionally in the speech of Persian speakers. The majority of informants avoided a direct refusal (mere no) and tended to provide reasons, explanations or excuses as a way to imply their lack of ability or unwillingness. This higher number of indirect refusals in participants' native language may be due to the cultural norms of Iranian society in which making a refusal directly even to someone of lower social status is considered as discourtesy. EFL learners, on the other hand, were more comfortable in using direct refusals.
Concerning the type of strategies, expectedly, 'reasoning' was the most common strategy in both groups (Eslami, 2010). EFL learners tended to use short structurally simple reasons. They employed a kind of "mitigated strategy". While Persian native speakers elaborated on the reasons and provided lengthier justifications for their refusals. Strategies 'consideration of addressee's feeling or opinion', 'statement of regret', 'letting interlocutor off the hook', and 'expression of gratitude' came to be the second to the fifth preferred strategies, respectively, in both groups.
There were six refusal formulas which were never used by the all participants in their Persians. These are direct performatives, statements of principle, statements of philosophy, requests for help, empathy, and assistance by dropping or holding the request, acceptance that functions as a refusal/lack of enthusiasm , and topic switch . The first of these refusal formulas is the most direct, and would consist of an utterance like "I refuse". In the case of direct "No" and "I can't", on the other hand, additional explanations for the refusal were commonly added. In the case of statements of principle and statements of philosophy, it would appear that using such refusals is seen as being overly philosophical in nature.
Frequency and content of refusal strategies between native Persian and advance EFL learners
Figure 2. Direct and indirect refusals and adjuncts to refusals produced by Persian native speakers (indicated by _P in the figure) and advance EFL learners (indicatd by _A) in four refusal situations in percentages
Advance learners used the same range of semantic formulas as native Persian speakers in their refusals but they differ in the frequency. Indirect refusal found the most trends among advance EFL learners as well. Like Persian native speakers, the advance EFL learners employed more indirect strategies in formal (farewell and advisor) than informal (birthday and notes) situations. However, in contrast to Persian native speakers, the advance learners used less indirect strategies and more direct ones in their speech. Moreover, the most frequent semantic formula was 'reasoning' in advance learners' speech.
In the solidarity (Birthday) and deference (Notes) situations, the common features characterizing these advance participants were the use of statements of regret. This formula was used in 23.3 percent of all the refusals. Unlike the case of excuses, statements of regret occurred at the beginning of the refusal in 71.4 percent of the cases. Furthermore, a statement of regret never occurred alone and was found at the end of the refusal in only 3 cases (10.7%). Statements of regret and excuses were usually closely linked; in 33 of 38 uses of statements of regret, an excuse either immediately preceded or followed it. In thirty of these cases (87%), the statement of regret came immediately before the explanation as illustrated in this response from subject 9: "bebakhshid, amma khodam laazemeshun daaram" (I'm sorry, I need them myself). In this case the statement of regret (I'm sorry) is immediately followed by a reason (I need them myself). This trend is accompanied by more avoidance strategies in native Persian responses.
Frequency and content of refusal strategies between native Persian and intermediate EFL learners
Figure 3. Direct and indirect refusals and adjuncts to refusals produced by Persian native speakers (indicated by _P in the figure) and intermediate EFL learners (indicatd by _I) in four refusal situations in percentages
The same rage of refusals strategies was witnessed in this category as well. However, the content varies. Intermediate group employed more indirect strategies than advance ones but less than native Persian speakers. The use of direct strategies among intermediate learners was less than advance group but more than native Persian. Just like advance learner content of speech, excuse got the most frequency in this category as well. Native Persian and intermediate learners both used reason but less than expressing regret. Often, first regrets were expressed, then reasons given. The Persian used regret more often than the intermediate. Generally the Persian used reasons after expressing regret when refusing strategies more than the intermediates.
Pause fillers (among Adjuncts to refusals) accounted for 8.1% of the refusal formulas among intermediate EFL learners. In the 22 refusals which used pause fillers, this formula occurred at the beginning of the refusal in every case (100%). Pause fillers were followed by a number of different semantic formulas. This variety of following formulas is not unexpected as pause fillers are typically used to give the speaker time to form their reply. The adjunct Gratitude/appreciation was used 6.3% of the time, while the adjuncts statement of positive feeling/ agreement and negative willingness /ability were used 6% and 5.3% of the time respectively. On the other hand, the rate of employing pause fillers was a little less than the intermediates (18 cases) among the native Persians. However, they enjoyed the same range of content as the intermediates.
Frequency and content of refusal strategies between native Persian and elementary EFL learners
Figure 4. Direct and indirect refusals and adjuncts to refusals produced by Persian native speakers (indicated by _P in the figure) and elementary EFL learners (indicatd by _E) in four refusal situations in percentages
This group had the most similarity among others to native Persian. Again, the content of semantic formulas was the same but the frequency was different. Indirect refusals enjoyed the most frequency in this category as well. The Persian speeches produced by native Persian and elementary learners had no significant differences regarding the content and frequency of the three categories of refusal