PRAGMATIC TRANSFER IN ARABIC LEARNERS OF ENGLISH

MOHAMMED GHAWI University o Arizona f

This study investigates socio-pragmatic transfer in Arabic learners of English. Following Olshtain (19831, the researcher attempted to assess the extent to which learners of English might transfer into L2 some of their first language (Ll)sod* pragmatic rules concerning apologizing. The closed role play instrument used by Olshtain (1983) to study the speech act of apology was employed. Findings suggest that L socio-pragmaticnorms were sometimes transferred to LZ. There 1 was also found some accummodation to L2 norms. The study also indicates that the extent of pragmatic transfer for certain apology strategies may be related to the learners'perception of the universality or the language-specificity o the speech act f of apology. Furthermore, the study contributes to the understanding of the possible generalizability of pragmatic transfer phenomena across different native and target languages.

INTRODUCTION
the role offirst Ianguage (L1 in second language (I21 learning. This renewed interest stems from )

There has been a revival of interest on the part of second language researchers in studying

the "overwheImingevidence that Ianguage transfer is indeed a real and central phenomenon that m s be considered in any account of the second language acquisition process" (Gass & Selinker, ut 1992, p. 7). In Selinker's m s recent model of the interlanguage hypothesis, language transfer is ot viewed as central to the development of interlanguage (Selinker, 1992). However, Selinker stresses the linguistic aspects of L1 influence (syntax, phonology, morphoIogy, lexicon). Socio-pragmatic competence has been under-represented in the early as well as recent models o the interlanguage hypothesis (for example, see Corder, 1978, Selinker & f Lamendella, 1981, Selinker, 1992) . Most of the research in the area of pragmatics has been influenced by Hymes' ( 1972) work on communicative mmpetence, wherein Hymes points out that language speakers acquire n t only grammatical competence, but also competence as to the o appropriateness of language use within the speech community,this being a kind of "tacitcultural knowledge "(Hymes,1972, p. 279). A basic research question in the last two decades has been to investigate how L1 socio-pragmatic competence affects the process of learning L2, hence the term pragrmlic t r m f e r appears. &per (1992) defines pragmatic transfer as "theinfluence exerted by learners' pragmatic knowledge of languages and cultures other than I 2 on their comprehension and production and learning o L2 pragmatic information"(Kasper, 1992, p. 207). She subsumes studies in this area f under a sub-discipline o second language acquisition (SLA) research known as interlanguage f pragmatics. In an earIier work, Kasper and DahI (1991) offered a narrower definition of interlanguage pragmatics as the study of "the nonnative speakers' (NNSs') comprehension and pduction of speech acts, and how their L2-related speech act howledge is acquired" (Ksper, 1991, p. 2 6 . 1) Blum-Kulka and Olshtain ( 19%)analyzed the utterance length of requesting strategies in Hebrew. Discourse Completion task data were collected from native speakers of Hebrew aad nonnative speakers from seven different languages at three proficiency levels. The researchers

Two Tidk, Vob 1, No. 1, Spring 1993, pp. 39-52.

were surprised to find that the utterances o high-intermediate subjects were longer than the f utterances of low-intermediate, advanced, and even native subjects. The verbosity of the high intermediate subjects was interpreted as a pragmatic failure because it violated Grice's (1975) maxim of quantity. In an earlier study, Blum-Kulka (1982) used the Discourse Completion also task to examine the use of the requesting strategies by American and Canadian learners o Hebrew f at intermediate and advanced levels. Her findings showed that the I 2 speaken developed an intdanguage for performance in this s p c h act t a w s different f m both L and I 2 norms. ht a I Trosbwg (1987) used role plays to investigte the apology stsategies of Danish learners of English at three proficiency levels. The performance o the subjects in L2 demonstrated fewer f explanations and minimizing strategies, which was attributed to transfer from L1. As their proficiency level in I 2 increased, the Danish subjects used more m d i t y markers to improve the pditems of their apoIogy strategies. Wolfson (1989) adopted a naturalistic approach for researching interlanguage pragmatics. She collected ohsemationaI data on compliments from authentic i n k d o n s between native and non-native speakers over a p r i d o two years. Her findings showed that L2 speakers did not f seem to understand the function o camphents as a social lubricant in the American culture. The f nm-native subjects alw had difficulty in respding appropriately l complhnmts. o In their 1991 article, Kasper and Dahl provided an extensive review of 39 studies i the n area ofinterlanguage pragmatics. The studies covered a wide range of languages, speech acts, and research immmnis. Some of the many independent variables included in those studies were a g ~ , sex, native language, proficiency level, length of stay, learner's perception, context, and developmental factors. The dependent variable was always related t the perfmmance strategies o o f speech acts. Kasper and Dahl were especially interested i assessing the validity of the research n instnunents used in the studies they reviewed. Ragmatic competence was also assessed by studies outside the scope o speech acts. For f example, Bartelt (1992)re-examid his 1 - findings o rhetorical transfer in Apachean English f based on subsequent research in the area of processing and nativization. In the original study, Apachean speakers transferred the rhetorid strategy of redundancy it Wir written cumpitions no in EngIish. As in LI,lexid and phrasal redundancy in I 2 was intended to serve the function of emphasis, especially for persuasion pupses. Looking at the phenomenon o transfer from a f pmedurdldeclarative prmwsing pmpct~ve, l t maintained that m f e r was attributed to 'n M a L procedural constraint in an L2 production systemw(Bartelt, 1992, p 103). He also add4 that 1 . this rhetorical transfer could be seen as part o a nativization paces o "cultural syncretism in f f which generic schemata act as constraints in selecting compatible features to fill in gaps in new howledge stmchms" (Bartelt, 1992, p. 1 7 . 0) Another example o research that investigated pragmatic competence outside the scope o f f speech acts was Scaroella's (1992)study o "discourse accent"in videotaped inter-ethnic f conversations. Her experimeatal group included ten proficient Spanish spealung subjscts. She found evidence of transfer i conversational features such as topic selection, back channel cues (for n instam, repetition of the interlmutor's previous utteranm], and pause fillers. In her intapretation of the findings, Scarceh provided two possibilities: (a) pragmatic fosdizatjon, (b) interlanguage evdving into a fully developed dialect.

THE PRESENT STUDY

The present study will use as a m a k l Olshtain's (19tB) m h to investi* the case of pragmatic transfer in L2 spoken discowse--specifidly, with regard to the apology s p c h act. The research questions to be developed parallel OlshWs:
1 To what exteat is there a tendency un the pari of the language learner to transfer socie . cultulal ruies from L to U? 1 2 Can such a tendency be predicted from the learner'sperception of language spcificity o . r language miversaIity in relatian t a certain s o c i d a u a I situation? o

3. How can these questions be investigated w t respect to the act of apologizing? ih (Olshtain, 1983, p. 233)

In addition, this study replicates Olshtain's ( m research in dl essential aspects in order 1 ) to contribute to understanding of the possible generalizability of pragmatic transfer phenomena
across different native and target languages.

Eight apology situations were first used in the 1981 Cohen-Olshtain study to collect data from native speakers o English and Hebrew that would establish acceptable norms of apology in f the two languages. These data were then compared with the responses elicited from Hebrew speakers learning English, using the same eight apology situations, as a measure of their communicativecompetence. In Olshtain's 1983 study the target language w s Hebrew; the two groups of subjects were a native speakers of English and native speakers ofRussian, both Iearners of Hebrew. The aim of the study was to gauge communicative competenm by comparing non-native responses to the same eight apology situations, used in the earlier study, with native speaker responses (Israeli speakers of Hebrew). As in Olshtain's study, the present study is designed to assess socio-pragmatic competence by describing non-native deviations in the specific socio-cultural context of apologizing, this being a significant element of the learner's overall communicative competence. Olshtain stipulated the need to address language transfer problems in terms of the interrelations among language-speci fic, cross-Iinguistic, and situation-specific features of a given type of speech act. Cohen and Olshtain ( 1981) p r o w the notion of a speech act set in order to account for such interrelations inherent in the speech act of apology. It was Austin (as cited by Olshtain, 19B) who first delineated the various ways a speech act can be performed using the appropriate perfornative verb (e.g., apologize),using another explicit verb (e.g., be s o y ) , or using a simple statement that indirectly performs the speech act within a specific context (e.g., an excuse o an r explanation). The apology speech act set served as the basis for cross-cultural: research and comparative data in the 1 8 Cohen-Olshtain and the 1983 Olshtain study, as well as the present 91 study. The use o one o more formulas for the s p c h act in question will depend more on sociof r cultural rules and the specific discourse situation, than on individual preferences ( 1 0 shtain, 1983, p. 235) &hen and Olshtain (198 1 acknowledged the possibility ofboth an acceptance and a denial ) of responsibility on the part of the perceived transgressor. In the case that the transgressor perceives the need to apologize, the following semantic fwmulas may potentially be utilized:

I. An expression of apology (I'm sorry) 2. An explanation or account of the situation (I'vek e n busy) 3. An acknowledgement of responsibility (It was my fault) 4. An offer o repair (Can I help you?) f 5. A promise of forbearance (It won't happen again)
Olshmn ( 1983) proposed a number of possible deviations that might m u r in L2 learners' performance of an apology as a result o inappropriate application o soci+cul turd rules: f f
1. The learner might deviate from the accepted norm when choosing a semantic formula f r o
a specific situation.

2. The learner might choose a combination of semantic formulas which is inappropriate for
a specific situation.

3. The learner might perform the speech act a a level of intensity inappropriate in relation to t
a particular offense. (Olshtain, 1983, p, 237)

The present study takes as its focus these same semantic formulas and potential deviations as applied to the case of speakers o Arabic learning English. f

Subjects
19921and Spring 1993 at the Center for English as a Second Language (CESL), located at the University of Arizona. A total of 1 Arabic speakers volunteered, 1 males and 1 female. Their 7 6 average age was 22 (range = 1&28 years). Concerning the educational levels of the participants, 12 graduated from high school, and 5 had at least two years at the university or technical college i n their home countries. Nine o the participants were from the United Arab Emintea, two from f

The L2 subjects were international students taking intermediate level courses in Spring

Qatar, two from Kuwait, and one each from Saudi Arabia, Morocco, and Oman, The average length of stay i the US for the group was six months (range = 2-12 months). The participants n had studied English as a Foreign Language in their home muntsies for at Ieast six years and most of them had traveled to foreign countries where English was the primary medium of

communication. Like the subjects in Olshtain's ( 1983)study, the subjects of this study were intentionally chosen from intemediate levels o L2 instruction. The choice of inmediatelevel students was f made because this seemed to be the most likely level of pwficienq at which one might expect both me of L SOCiopragmaticrules and accommorhlion Eo L2 rules. fr 1 In addition, a control group of 17 native English speakers participated in the study. Members of the control group were 12 females and 5 males, all of whom were undergraduks a t the University o Arizona during Spnng 1992and Spring 1993. Their average age was 21 (range = f 18-22years).
Instrument As in Olshtain (1983), the basic instrument used for collecting data on the speech act of aplogizing was c l o d role plays. This data collection technique involves audio-taping role play situations between the researcher and the participant The eight roie plays involving apology were performed both in L1 (by native Arabic-speaking and native bghsh-s-ng subjects) and in L2 (by native Arabic-spwkng subjects d y } (see Append~x Another instrument usad, after both A). L and L2 role plays were completed, was a short intetview with each Arabic participant for 1 collecting personal data and information on the participant's perception of the specificity or univessality o apology across languages [see Appendix B. This intwview was conducted in L f ) 1 fm the purpose of collecting as much valid infomalion as psible from the subjects.
Procedure Each participant was given a written description o each role play situation on an index f card, so that he or she could read the d b p t i o n and come to an understanding o what was being f asked o his or her performance. The first =ssion with each prticipnt involved performing the f role play situations in the first language. One the participant signalled his or her readiness to begin the role playing, the tape recorder was used to m r d tfie dialogue. The same p d u r e was used for each situlion and in both L1 and L2 sessions. A week: i n t e d was wed between the two sessions. All the role plays with the Arabic spaken,both in L1 and L2,were conducted with the mearcher, a native speaker of Arabic. In addition, eight of the role plays with the English speaking control group were conducted with the researcher himself, whereas the other nine were conducted by two native speakers who were teachers o the two classes fmm which the c01ltroI f group was chosen. Both teachers were familiar with the instru~lla After the role play shmtions were d e d in L2, the partitipant was interviewedt collect o ma n data and was questioned abut his o her perception o apology, as noted above. The two l r f questions used to elicit this information are
1. Do you think that speakers o English apologize more o less than speakers of yaw f r native language? 2 Do you feel that a native speaker of English might apologize differently from a speaker of your language for any of the eight situations? (Based on Olshtah, 1 8 ,p. 239) 93

md T~ferinAd+icLeam?rn English 4 3 o f
The two questions were orally translated by the marcher into Arabic in order to make sure that the pwbcipants dearly understood the questions. They were given as much time as they needed to write down the answers in their native language. A recorded oral discussion o the written f answers followed between the participant and the researcher. If a Wcipant said that English speakers apologized differently from speakers of their native language, this response was interpreted as registering a language specific perception of the speech act o apology. If a f parhcipant stated that it was the situation which determined how they apologized, not the Ianguage, then this -me was interpreted as reprmting a universal perception
Data Analysis The researcher transcribsd the tape recordings and then coded for the five apology stmtegies, M h g for the use of a strategy was h a d on presence or absence of the strategy. Therefore, the data reported in the following analysis will be based on the presence o absence of r

strategy coding only. After the d i n g was completed for the native English (NE), native Arabic (NA), and Arabic i English (AE) n group, comparisons were made between the responses of the three groups. In addition, the responses to the two interview questions on a participant's perception of apologizing were also used in interpreting the data. For instance, if a mcipant's use of apology strategies incorporated any L sociopragmatic rules, then, in accordance with 1 OIshtainfsprevious findings, the present researcher predicted such a participant would have a

m i v d perception o apdogy, regardlas of the language used. f

RESULTS

Situation #1: Insulting someone at a meeting The data in Table #1 show that the three apology strategies used by speakers o the three f language group were: direct apol~gy, explanation, responsibility. Repair and foheamce didn't seem to be appropriate for apologizing t a person who felt insulted at a meeting. Only one native o speaker offered to repair saying, nI'dlike to repeat myself and explain it more c l d y . " Native English (NE) speakers demonstrated a higher o apoIogy and responsibility than f native speakers of Arabic (NA). The native English nonn was 9 % for apology and 82% for 4 responsibiIity, whereas the native norm for Arabic speakers was 71% for apology and 76% for responsibiiity. Arabic speakers felt a lesser need to express apology or acknowledge responsibility.For example, some explained that in a meeting, individuals should not feel insulted. In addition, their use of explanation o the offense occurred a a higher rate than the explanation f t strategy of the NE speakers (NA = 7 %; NE 29%). As one Arabic s@er put it, "Mytalk was 1 general in nature. If you thought you were insulted, this indicates that you were guilty of something."
Table #1 Percentage of Each Strategy for Situation #I

NE = Native English, n = 17; NA = Native Arabic, n = 17;AE = Arabic in English, n = 17

The Arabic speakers' performance in I 2 was identical to the L norm in the use o the I f apology strategy (NA = 71; AE = 7 )and a little bigher thaa the L norm in the use o explanation 1 1 f (NA = 71;AE = 76).However, their rate of use o the responsibility strategy was far less than f both NE and AE norms. The similarity i strategy use between L aad L2 for apology and n 1

explanation might be attributed to L1 influence, whereas the dissimilarity of the responsibility strategy use from ~ I I L and I 2norms could be due to developmental factors in the interlanguage I1 pragmatics of the learners.
Situations #2, 3, and 4: Forgetting a meeting with the boss; forgetting a meeting with a Mend; forgetting a meeting with your son These three situations were designed to formulate a continuum of formality in the relatiomhip between the person apologizing and the person apologized to. To interpret the data in situations 2-3, and 4, I compared the native data within each native language and then across the ~ v native languages before studying the ssimiIarities and differences between native and non-native o responses (see Table a).

Table #2 Percentage of Each Strategy for Situations 2+3+4

NE

NA

AE
82

Situation #2-Forgetting a meeting with the boss: APO~OW 100 76 Explanation 76 88 Responsibility 7 6 24 47 18 Forbamme 29 18 Situation #3-Forgetting a meeting w-ith a friend: AP~~WY 94 76 65 76 Explanation 65 41 R e s p s i bility 47 24 Repair m c e 24 12 Situation #&Forgetting to take your son shopping: APO~Q~ 100 59 53 7 1 Explanation 1 8 35 Responsibility 84 88 Repair

10 0

m=

53
29 12

82
82
6 82 47
35

41

24

Fmbamlm

18

88 1 2

NE = Native English, n = 1 7 NA = Native Arabic, n = 1 7 f f i = Arabic in English, n = 1 7

For NE speakers, the data show that they maintained a very high percentage o apology f across the three situations (100%; 94%; 100%). Only one NE speaker failed to use a direct apology strategy when apologizing to a friend. Instead, she explained the reason and acknowledged her responsibility by saying, T have a lot going an right now. I just totally forgot about the meeting I had with you."Theother strategies were much less frequently used exapt for repair to a child (94%). The use o both explanation and responsibility strategies gradually f i n m a d with the increase i the level o formality (53%,18% to a child; 65%,65% to a friend; n f 76%, 76% to a boss). NA speakers used fewer apology strategies than NE s@m in the three situations. Their apology strategy was the same to a boss and to a friend, but was much less when apologizing to a son (76%;76%;59%). Their use o the qmnsibility strategy was also lower than the NE group f in two of the three situations (24%to a b; t a friend). As in Situation #l, explanation 41% o seemed to be a very camonly used strategy i Arabic. The use of this seemingly basic apology n strategy in Arabic also increased with the increase i the level d formality (71%to a son; 76% to a n

friend; 88% '0 a boss). As i L1 English, repair and forbearance were not frequently used except n for repair .to a child (88%). The performance of the Arabic speakers in English showed similarity to native Arabic norms, accommodation t native hglish norms, and sometimes dissimilarity from the two native o languages. For example, their use of apology strategies in Situations R+3 (82%; 82%) were 6% hi@w than L1,but 18% and 12% lower than L2,indicating similarity to L . Furthemore, use of I repair and respnsitili ty in Situations #2+3 were at percentages identical to those o L1(41%, 35% f far responsibilty; and 24%, 88%for repair, respectively). However, when apologizing and explaining to a son, the AE group demonstrated considerable accommodation to L2 norms (82% for apology and 47% for explanation). Like the norms of the two native languages, Arabic and English use of explanation and responsibility strategies once more increased with the increase along the formality continuum (47%- 35% to a son; a%, to a friend; 100%, 53% to a b s , respectively). Interestingly, 41% not only did the AE speakers maintain high use of explanation strategies in Situations #2+3, but also they increased them t 100% in Situation #2, which was 12% higher than in L . o 1
Situation #5: Backing into someoneis car and causing damage This situation, as i OIshtain (I=), was incIuded to provide an appropriate context for the n repair strategy. The data i Table #3 show that the native speaker norms for repair in the two n languages were the same (82%),whereas the use of responsibility was higher for NA speakers (NE= 65%; NA = 82%).However, Arabic speakem demonstrated less propensity to make both offers of repair and achowledgement of responsibility when speaking i L2 (65%for repair; 59% n f r responsibility). Expressions of apology for Arabic speakers did show an increase from 53% in o L to 94% in U, 1 which put them very close to the target language of 100% for this situation. Table #3 Percentage of Each Strategy for Situation #5

Apology Explanation

100
24

Responsibility

53 24
82

94
35 59 65

ReParr

Forbeamice

65 82 0

82 0

NE = Native English, n = 17 NA = Native Arabic, n = 17 AE = Arabic i English, n = 1 n 7

Situations ##, 7, and 8: bumping hto a lady and hurting her; bumping into a lady and shaking her up a bit; bumping into a lady-her fault These three situations were related, and were meant to elicit a continuum of intensity of regret dependent w the gravity o the offense, with Situation #6 constituting the most serious f infraction, Situation #7 a l e ~ serious offense, and Situation #8 requiring perhaps no apology (see s Table #41. ~ative speakers of English were fairly consistent in expressing apology for each of the thee situations (94%;100%; 100%). Only one NE speaker failed to use a direct apology fonn. Instead, she chose to inquire about the well-being of the lady and offered to help her saying, "Oh, are y w O.K.? L t me help you pick up your bags. Do you want me to help you? Are you e 0-K?"The use o the responsibility stratew seemed to increase as the severity o the offense f f increased (29%; 41%; 65%). Offers of repair were also high for Situation #6, which involved the strongest offense (76%),mmpared to Situations #7+8 (6%;128, respectively) which involved less serious offenses.

Native Arabic speakers demonsmd a fairly high 'and consistent use of the apology strategy ( 1 W ; 88%; 100%) across the three situations. The two subjects who did not offer direct apology seemed offended by the lady's relatively rude reaction to the offense "Hey, look out!" One of them responded sarcastically, "I should have been more d u l , but your beauty distracted my attention," whereas the second responded anjjrily, "Okay, Okay. I d i d n h e you. How many
eyes do I have?"

Table #4 Percentage of Each Strategy for Situations #6+7+8

NE

NA

AE

Situation #6- Bumping into a lady and hurting her: Apology 94 100 100 Errplamian 0 6 12 Responsibility 65 71 65 76 76 7 6 ReParr Fr o0 0 0 Situation #7-Bumping into a lady and shaking her up a bit: APQlW 100 88 100 Explanation 24 29 18 Rqonsibility 41 82 65 Repair 6 6 6 ForbearaTlcle 0 0 6 Situation #&Bumping into a lady-her fault: Apolw 100 100 100 l%pl&m 12 71 71 Rqmnsibility 29 65 35 12 6 1 2 Repair Fabearme 0 0 6

NE = Native English, n = 1 7 NA = Native Arabic, n = 1 7 AE = Arabic in English, n = 1 7

The use of the repair strategy by the Arabic s p k m seemed t have k e n influenced by the a gravity o the offense. It was considerably high in Situatian #6 (76%) compared to Situations f # 7 4 (6% f r each one). Acknowledgement o mpmibility did not appear to be influenced by o f the severity o the offense continuum. Actually, Situation f7, which involved less offense than f Situation #6, elicited more responsibility strategies (82%;7 8 for Situations #7+6 respdively). 1 Explanation, which appeared to be a basic apoIogy strategy i the first five situations, was m l y n used i Situations #6+7 (6%;29%, respectively), It was in Situation #8 when explanation was n used again by the NA speakers, probably because these subjects had to explain to the lady that she was at fault--standing in the way. As one of the subjects said, "Swry...sorryaunt. I swear to God, there is no other way. I tried to avoid you, but 1 wuldnft. " When using English, all Arabic speakers used the apology strategy i all three situations. n In fact, their rate of use of this strategy was higher than both NE and NA speakers. In Situation W8, involving the least responsibility on the part of the apoIogizer, Arabic speakers showed a decrease in offem o responsibility from 65% i NA to 35% i AE--nearfy parallel to the NE norm f n n o 2 % A similar, but lesser, demease in the use of mpnsibility tmk place in Situation #7 from f 9. 82% in NA t 65% in NE. This d o e did not bring the AE subjects clae t the NE norm o o f 41%. However, the use o explanation as an apology strategy in AE was identical to the NA norm f (71%) and much higher than the NE norm of only 12%.

Average Use of Strategies Across the Eight Situations: To provide a clearer picture of the similarities and differences among the three language groups i the use of the five strategies, 1will follow Olshtain's (1983) example of summarizing the n average use ofeach strategy across the eight situations. When comparing the o v d l frequency of strategy use across the eight situations for the three language groups (see Table #5), one finds that the native English speaken tended to use the direct apology strategy with more frequency than Arabic speaken in both L and L2 (NE= 98%; 1 NA = 75%; AE = 89). They also had a slightly higher rate of use of the repair strategy than the other two p u p s (NE= 4 8 NA = 38%; AE = 38%). Arabic speakers, on the other hand, used 6 ; more explanation strategies, probably in compensation for the fewer offers of apology and repair, when they spoke i both L1 and LZ ( I = 35%; NA = 55%; A E = 55%). Two of the L2 apology n FE strategies of the Ambit s e e r s , ercplanation and repair, were identical t those in L1(55% for o explanation;38%for repair), whereas one of the slmkgies, respnsibility, was Iower than both L1 and I 2 (NE=S5%;NA = 60%;AE51%). Table #5 Average Frequency (%) of Strategies Across the 8 Situations

AE = Arabic in English, n = 1 7

NE = Native English, n = 17 NA = Native Arabic, n = 1 7

To further ascertain the relations holding between strategy use and language group, the Y ates Corrected Chi-Square test was used (see Table a . )
Table #6 Yates Corrected Chi-Square Results A m s s the 8 Situations

Strategy ( m g . Grps.1

Chi-square
Value

DF

Prob.

Apology (NE*NA) ANogY ( W A Q l%planation (NE*NA) Explanation ( W A E ) Raponsibility (NE*NA) Responsibility (NEYAE) Repair ( W N A ) Repair W A Q Fmbrance (NE*NA) Forkarance (NE*AE)

There is no relation between language group and use of sbakgy. Null Hypothesis: Alternative Hypothesis: There is a relation between languge group and we of strategy. Dependent Variable: Apdogy Strategy (presenceand absence). Independent Variable: Language Group (NE;NA;AE) "Significant at p<05

The results o this statistical prwedm suggest the following: f

(1) Use o the apology strategies Direct Apology and Explanation in L1 English and in L f 1 Arabic are statistically signifiwntly different (Alkmtive Hypothesis q t e d ) . (2)Use of the apology strategies Direct ApaIogy and Explanation in L English and i L2 1 n English are statistidy sipficandy different (Alternative Hypthesis - . 1 (3) U e of the *logy strategies Responsibility,Repair, and Forbeaatlce i L1 English s n and in L Arabic are not statistidly simmntly different (Failure to reject Null Hypothesis). 1 (4) Use of the apology strategies Respwsibility, Repair, and Forbearance in L1 English and i LZ English are not statistically significantly different (Failure to reject Null Hypothesis). n

DISCUSSION
The examination o the speech act of apology through the use of closed role piays has f provided some interesting insights into the extent of pragmatic M e r i the interlanguage of these n f Arabic learners o English The data from the role plays and short interviews has helped to assess the extent to wbich pragmatic transfer might be related to the learnerst perception o language specificity or f u n i v d t y . The native Arabic speakers i this study all remarked i their interviews that they felt n n Americans apologized differently from Arabs, specifically that Americans apologized mare frequently and at times unnecessarily. For instance, some of the Arab partrcipants stated that Americans even apologized to their children, implying that h s was less common i Arabic. Many n of the Arab participants felt that although Americans apologized more, they actually meant it Im; that is, t apologies were less sincere, k The Arab participants all felt that the stmtegies of apology were dependent on language specificity. Olshtain suggests that the more the speaker perceives apology as being language specific, the more the speaker will be able to accommodate to the pra-c norms o the second f language. The data i Table #5 show that the Arab participants seemed t accommdate in the n o direct apology strategy only, with a 14% increase in strategy usage toward the native English norm. Still, Table #6 shows that the use of the direct apology strategy in L2 English was nevertheless significantly different from L English norms. l The Arabic I m e r s ' prception of Ianguage specificity did not seem to have influend their use o the strategy was identical t that of f o them in their use of the explanation strategy L. J Arabic (55%),which proved to be si@cantly different from L English norms (seeTabIe 1 #) Interestingly, the frequencies for the mpomibility shategy in L2 were lower than those of 6. both native languages (NE = 55%;NA = 60%; AE = 51%) ,which indica&d that factors other than either L1 o L2 might influence the development o interlanguage pqmatics. In comparing the use r f of the apology strategies Responsibility, Repair, and Forkmnce in L1 English with b t h L 1 Arabic and I 2 English it was found that there was no statistically significant dffmnce (see Table #6). CONCLUSIONS

To my knowledge, there has not been any research involving the speech act of apology in native Arabic or Arabic speakers in English. This study suggests that pragmatic transfer phenomena may be generalized across different native and target languages. In addition, the results o this study might have same important culturd and pedagogical implications. Since f miscommunicationsometimes occurs between native speakers of the same language, it may n t be o unreasonable to assume that miscommunication is likely to take place even more of& i n intercultural communication. One example from this study is the Arabic speak& interpretatiun of the frequent use of the direct apology strategy by native English speakers. Dwing the short interview sessions that followed the role plays, the majority of the Arabic subjects wmk on the interview form o told me orally that % n y k r "excuseme" in English does not mean much to r them in terms of the sincerity of apology. It seems to me that most o them understod (or f

probably I should say m i s u n d d ) this basic apology strategy in English, as can be seen from the mu1ts of this study, as a meaningless routine. Similarly, during an o a presentation of the rl preliminary results of this study in one of my classes, I was surprised to hear similar misunderstandings on the part of some native English speaken with respect to Arabic usage. At Ieast three members of the audience commented that the frequency of using the explanation strategy by Arabic speakers, which appears to be a basic apology strategy in the Arabic culture, is an avoidance tactic. As for the pedagogical implications of the study, the researcher would like to stress the signiF1cance of integrating culture into L2 instruction. Lack of sociqnagmatic teaching and error correction d d have been the reason for what Scarcella (1992) called a fossilized "discourse accent"iinadvanced L2 speakers of English. O m a g o (1986) reminds us that in addition to being acquainted with the linguistic aspect o a language, the L2 learner should have knowledge o "the f f patterns of living, acting, reacting, seeing, and explaining the world of the target country as well" (p. 359). If communicative competence is a major goal o L2 learning, leamew should not only f h o w the forms ofL2, also they should understand the socio-cuttuaI contexts for using them. but

ACKNOWLEDGMENTS
I would like to express my thanlrs and appreciation to D .MurieI Saville-Tmike, D .Donna M r r . J o h n , and Dr. Douglas Adamson fortheir invaluable comments on drafts o this paper. Special f thanks are due to my friend Phillip Elliott for his help with statistics and for his generous support throughout my work on this paper. A specid thanks is also extended to Rcd Tyson for his insightful comments on an early version of this paper. To conclude, I would like to sincerely tbank all the students who participated in this study.

THE AUTHOR
Mohammed Ghawi is a graduate student in the Interdisciplinary PhD. Program in Second Language Acquisition and Teaching, University of Arizona. His major is LZ Pedagogy and 'Program Administration and his minor is L2 Prooesses. H has BA and Diploma degrees in e English from Aleppo University, Syria. H also has an MA degree in D L from Saint MichaeI's e College, Vermont. He is currently teaching ESL at the Center for English as a Second Language, University o Arizona. He has taught w i s h in Syria for six years, two years at the university f level and four years in high school. His main research interests include interlanguage pragmatics, cognitive styles, and second language attrition.
NOTES lResearch conducted in Spring 1% was part o a larger m c h project on the same topic f involving native Spanish spealang learners of English as well as the native Arabic speaking learners mentioned here. Besides the author, the other researchers of the larger study included Shelley Omahony and Phillip Elliott. REFERENCES

Bartelt, G. (1992). Rhetorical transfer in Apachean English. In S.M. Gass & L Selinker (Eds.), . Lmguuge transfer in language learning (pp. 101-108). Amsterdam: John Benjmins Publishing Company, Blum-Kulka, S.( 9 2 . Learning how to say what you mean in a seoond language: A study of 18) speech act performance of learners of Hebrew as a second language. Applied Linguissics, 3, 29-59. Blum-Kulka, S. & Olshtain, E (1986). Too many words: Length of utterance and pragmatic failure. J o u m l o Prapatm, 8, 47-61. f

Cohen, A. & Olshtain,E. (1981). Developing a measure o sociOCUIturaI competence: The case of f apology. Language Uatning, 31,113134. Corder, S.P. (19'78). Language-lmer language. In J. Richards ( d ) Undersbatnding second E., andforeign lunguage leanring. Rowley, MA: Newbury House, Gass, S.M. & Selinker, L (Eds.) (1992).Language Zranrfer in laragrcage kmnhg. Amsterdam: . John Benjamins Publishing Co Grice, H.P. ( 9 5 . Logic and conversation. In P. Cole & J. Morgan (Eds.), Syntax and 17) senuzdcs: Speech EECrs (mi.3). New York: Academic A s . es Hymes, D. (1972). On communicative competence. In J.B. Pride & J. Holmes (Eds.), S~ciolirnguistics~ Harmwdsworth, Middlesex: Penguin Boob. Kasper, G.( m . 1 )Pragmatic transfer. Second Lrsngtaage Research, 8 ( ) 203-23 3, 1. Kasper, G.& Dm,M. (191). Research methods in interlanguage pramti=. SSLA,13,215
247. Qlshtain, E.(1983). Socio-culW wmpeknce and language transfw. The case o apology. In S. f Gass & L SeIinker . Language trunsfer in language karning (pp. 232249). Rowfey, Massachusetts: Newbury Howe Publishers. Omaggio, A.C. ( 9 6 .Teaching langmge i conmt. Boston, MA: Henle & Henle Publishers, 18) n

m.),

Inc*
S-lla,
R (1992). Interethnic conversationand second language acquisitim D s o r e accent ious revisited In S M. . Gass & L Selinker (Eds.), Langmsage Transfer i h g w g e karning . n (pp. 109137). Amsterdam: John Bmj& Publishing Company. Sinker, L (1992). Rediscovering iptterlangmge. London: Longman. . Selinker, L. & Lamendella, J.T. (1981). Updating the interlanguage hypothesis. Studies in Second LQngwge Acquisition, 3 ( ) 201-220, Z, Trosberg, A. (1987). Apology strategies i nativdnon-natives. J o w d of Pra@s, n 1 1,14767. Wolfson, N. (1989). The social dynamics of native and nonnative variation in complimentary behavior. In M.Eisentein (=),?Re dynamic interlanguage (pp. 219-236). New York Plenum. APPENDIX A Apology Instrument Instructions You will be asked t read eight brief situations calling for an apology. I will role play this person. o Respond as much as possible as you would in an actual situation. Your r e s m will be taperecorded, Indicate when you've finished reading. (The following situations were presented on cards in random order.)
Situation 1

You are at a meeting and you say something that m e o the participants interprets as a personal f insult to him. H :"I feel that your last remark was h t e d a me and I take offense." e t

You:

Situation 2

You completely forget a crucial meeting at the offim with your boss. An hour later you d l him t o aplogira. The problem is that this is the second time y&e f r o t n such a meeting. Your boss ogte gets on the line and asks: Friend: "What happened t you?" o You:

Situation 3

You forget a get-togetherwith a friend. You d l bim to apologize. This is already the second time you've forgotten s c a meeting. Your friend asks over the phone: uh Rend: "Whathappened t you?" o Ywr:
Situation 4 You call from work to find out how things are at home and your son reminds you that you forgot to take him shopping, a you had promised, and this is the second time that this has happened. s Your son says over the phone: Son: "Oh, you forgot again and you promised!" You: Situation 5

Backing out of a parking place, you nm into the side of another cat. It was clearly your fault. You dent in the side door slightly. The driver gets out and comes over to you angrily. Driver: *Can't you look where you're going? See what you've done!" You:
Situation 6

You accidently bump i t a well-dressed elderly lady at an elegant dqmhnent store causing her to no spill her packages a l l over the floor. You hurt her leg, too. It's clearly your fault and you want to aplogize profusely.

Ym

She: " !

M y

Situation 7

You bump into an elderly lady at a department store, shaking her up a bit. It's your fault, and you want to apologize. She: Wey, look out!"

Yw:

Situation 8

You bump into an elderly lady at a department store. You hardly could have avoided doing so kcawe she was Mocking the way. Still, you feel that some kind o apology is in order. f She: "Oh,my!" You:

APPENDIX B

Short Interview Form

Name:
AF MaleFeMaIe: comtly:
Native Language: E d d m Levd: Class Level a CESL t Languages Spoken: Length of Stay in US: Travd Expien=

Learners1 Perception of Apology

1 Do you think that speakers of English apologize more o less than speakers of your native . r language?

2. Do you feel that a d v e speaker o Eaglish might apologize differently from a speaker ofyour f language for any of the eight situations?

RESPONSE TO GHAWI
DONNA M.JOHNSON Director, Englirh isAnguugelLingu5ticsProgram Univetsiby o Arizonu f
Mohammed Ghawi's study m k s an interesting contribution to the growing research ae literature in cross-cultural pragmatics and provides an impetus for further research on AmbidEnglish pragmatic contrasts. The study essentially addresses three questions: (I) How are apdogies realized differently in Arabic and in American English? (2) How do Afabic speakers apologize when using English as an additional language? and (3) What are Arabic and English speakers' perceptions o cultural differences in apIogizing? The simultaneous focus on these three f q d m s - - t h e mmparative question, the interlanguage question, and the perceptual question--gives the study a certain multidimensional interest. Ghawi has not only u n c o v d some important crascultural contrasts in apologizing, but has begun to draw connections between Arabic speakers' prqtions of pragmatic differencesand heir p e r f o m = in M Arabic and English. A key comparative finding of the study is that, given comparable situations, Arabic speakers were less I W y than English to offer an explicit statement of apoiogy (e.g., "I'm sorry") a part of their response. Arabic speakers reporred that h e r i m s apologize too frequently, sometimes unnecessarily, and less sincerely than Arabic speakers. When Arabic speakers apologized, they were more likely to o f r explanations for the offense than were English speakers. fe Ghawi reports that some English speakers view this strategy as an avoidance strategy. While the perceptual component o Ghawi's study is very limited, the examination of perceptual information f and discourse production within the same study can help explain reasons for cross-cultural misunderstandings. Data were elicited in using eight role-play situations that had been developed and used in previous research. These data were supplemented by a short questionnaire and by interviews with the subjects about their pmphons of apologizing i each language, n Perhaps the most valuable aspect ofthe study i the comparison between native s p k e r s of s Arabic performing in Arabic and native speakers of English performing in English. Ghawi established a set of (essentialIy quantitative) baseline "normsu (using the term loosely) for performance of these discourse elicitation tasks in a s p i f ~ research situation for both p u p s . c These *normsnprovide a basis for examining the performance of the Arabic speakers in English, which is their second (or third, ew.) language. Z would also be desirable to examine the t performance of native +IS of English leaming Arabic, but few such students were available. Many studies o problems in intercultural communicah examine only the performance of Ieamers f o a language, wmpring their language use with that o native speakers of that language. While f f such studies can provide valuable infomation about discourse productions (e.g., ScoIlon, 1993; Tyler, Jeffries, & Davies, 19881, there is little basis for drawing conclusionsabout the reasons for the pattern observed, including possible L1 transfer. Ghawi's study avoids this problem by examining Arabic speaker's L1 performance as well as their performance in English. His study, then, falls into the categories of (1) comparative studies, as well as (2) studies that examine interlanguage m a t i c s in the context of native Ianguage use in both of the languages of interest. Yet, there are a number o validity questions that mwt be raised. First, how well does the f Arabic speakers' performance on the discourse elicitation task represent their actual Arabic language use in everyday interaction with other Arabic speakers? The best way to answer this question, and thereby conduct a validity study, is to gather extensive naturalistic discourse. This approach would be quite time consuming because apologies may not occur frequently. A compromise might xto draw examples from existing sources such as Iiterature, o to devise r creative, semi-natwalistic data mIlection techniques. Second, how well dces the Arabic speakers' performance reflect their language use in Arab cultural contexts?That is, how relevant are the task
QTwo Tdk, VoL I , No. I , Sp*1993,

pp. 53-55.

situations to life in the Arab world? There may be some important cultural differences in the typical situations that call for an apology. A more cultudly-oriented comparative study might employ, if not ethnographic techniques, task situations developd in the cwtext o each culture. What is lost f in comparability and control from an eqmimatal peqective w d d be offset by what is gained i n cultural authenticity and cultural expImation. Now let us consider some validity questions regarding the performance of the Arabic speakers i English. First, it would be interesting to examine the English performance o Arabicn f speaking students of English at campamble levels o proficiency who had not lived in an EngIishf speaZring counby. One might expect patterns of perfomme in English that would be more similar to those of Arabic speakers using Ambic. Second, it is important to consider how the ethnicity, language abilities, and other characteristicso the rsamher conducting the role plays might have f affected the performances o the subjects. For example, how might the subjects have performed f differently if they had enacted the English roleplay situations with a native speaker of English from the U.. Betxuse all of the Arabic-spealung subjects expressed the view that one apdogiz~s S? differently in Arabic and in Enghsh, it is possible that they made a strong attempt to use what they perceived as a U.S. apology style. Oa the other hand, some subjects might have used (either consciously or u n . ~ c i o u s l y Arabic ways of apologizing because the researcher was a native ) speaker o Arabic. These issues compIicate interpretation o the data but they are also worthy of f f further study; that is, one muid examine in more detail advanced students' ability to vary their pragmatic or discourse style acmrding to a variety o situational facton, including characteristics of f interlocutors.

Perceptions A most interesting, although very limited, part of this study is the examination of the subjects' perceptions o the "languagespecificityQr the "universalitywf apologizing. By briefly f o interviewing each of the subjects, Ghawi Ieatned hat all o the s p k e r s of Arabic reported that f Engiish speakers apologize differently. WhiIe this might have been the answer they h e w he was expecting, there i some consistency amang these reports, their further comments in the s interviews, and their discourse performance data. For example, in interviews, Arabic-speaking subjects reported an awareness that Ameicans apologiz to their children to a greater extent than in Arab cultures. The data bear this out. While 100% o the Ameriwns provided an explicit apology f to a child and 53% provided an explanation, among the Arabic-swng subjects, only 59% offered an explicit apology to a child, while 71% offered one (ormore) explanations. When using English, this group of Arabic speakers exhibited patterns o performance in the use of these two f strategies that fell between the Arabic and English %oms." . GhawiL analysis of the language specificity or universality of apologizing involved dichotomous d n g ; that is, students answers were coded as either reflecting a language-specific perception o apologizing o a universal prcepEion. Because the m p e s of all subjects were f r categorized as reflecting a language-specific view of aplogizing, the most interesting information fmxn these brief questionnaim and interviews involved the specific kinds o language differences f the subjects perceived and how they r e W these differences to cultural behavior and v a l w . Yet, this information was not reported in a systematic way. There is much mom for further m h in this area, specifically for innovative approaches to linking language performance to cultural perceptions. For example, extensive interviewing involving systematic analysis o the qualitative f data would be praductive as would e@mmtaI studies using v i d w t a p i roleplay enactments to compare pmeptions o language use and c u l W appmphtenm (see Kasper & Dahl, 1 9 ,for a f 91 discussion of methddogical problems i rnehpragmaticjudgment studies). n

Analysis Other issues regarding analysis of the data raise questions. Because functional d i n g of discourse data is not straightfornard, interrater reliability should be established based on definitions developed i the m t e x t of each @fit study, n Speech act data were ccded for the pmentx o absence o a strategy. Frequency of use data r f were not reported, however. Are there any salient frequency findings that would enrich the data

Responre i Ghawi 55 b

that are ~ p t d For example, did the s p k e r s of Arabic tend to give tong ql&ons in Arabic ? and shorter expIanations when they were using English? What was the nature of their explanations? Did the native English speakers tend to use one semantic formula for expressing an explicit apology or two? What were these? Information such as this would enrich the comparisons.
Discourse This study is abut language use but, because of the quantitative orientation o the study, f we see very little of the actual language that the subjects produced This is a problem that is typical of much o the research in cross-cultural and interlanguage pragmatics, while research in the f Gumperz tradition, for example, emphasizes discourse (Gumpen, 1592). Authors of studies i n cross-cultural and interlanguage pragmatics should provide typical examples of discourse that illustrate some of the key fmdings of a study. For example, it would be most interesting to readem if Ghawi would provide several typical comparative examples of complete Arabic L responses 1 and English L respses to situations that illustrate, for example, how the Americans provided an 1 explicit apology statement, while the Arabic speakers emphasized explanation. Providing rich examples o actual discourse would not only enhance the &cle as a report o research, but it f f would make the finding more useful to materials developers, teacher lmhers, and learners.

Interpretation An important goal of the study was to find evidence of transfer of pragmatic strategies from Arabic to English. The fact that percentages of students using particular strategies i n intedanguage performance ofen fed between percentages o students using these strategies in their f Lls provides some indication that transfer f o the L might be a factor. However, other factors rm 1 affecting prfmance must be considered as w l ,such as those discussed above. If some of the el students were mu1tilingnal, it would be important to consider how knowledge of other languages and cultures might have influenced their performance and perceptions. Also, in some cases a student may have wanted to make an offer o repair, let us say, but may have been unsure of the f correct way to express the offer and may have omitted it. I would like to see more of the author's own cultural interpretation of the meaning o the f differencs he found in m a t i c strategies. For example, some subjats expIained that "in a meeting individuttls should not feel insultedn(p. 4) This sbtement requires further explanation 3. for non-Arabs. Because the author is a native speaker o Arabic who is highly bicultural, readers f would be interested in his interpretations of the reasons for some of the phenomena he obsemed, even though his explanations might be speculative. His own interpretations would be part~cularly valuable if he compared them with inmpreta60ns o other native qakers o m c . f f In conclusion, work in cross-cultural and interlanguage pragmatics can be very important not only in addressing issues o comparison and transfer, but d timely in promoting better m f cultural understanding. However,t better achieve these goals, a stronger emphasis on naturalistic o o ethnographic data mllwtion and on the analysis of discourse in context should lead to richer r insights. REFERENCES
Gumpen, J.J.(1992). Contextualization and understanding. In A. Duranti & C. Goodwin (Eds.) , Rethinking context: Lunguge as an interactive phenomenon (pp.229-252). Cambridge University Press. Kasper, G. & Dahl, M.( 1 9 ) Research methods in interlanguage pragmatics. Studies in Second 91. Language Acquisilion, 1 3 , 215247. Scollon, R. 11993). Cumulative ambiguity: Conjunctions in Chinese-English interculturd communication. Perspectives: Working papers of the Department of English, City Polytechnic of Hong Kong, 5(I), 55-73. Tyler, A. E ,Jefferies, A. A. & Davies, C.E.(1988). The effect of discourse structuring devices . on listener perceptions of coherence in non-nati ve university teacher's spoken dr scourse. World Englishes, 7(2), 101-110.