* Conventions of Malaysian Grocery Store Service Encounters

The study aims to describe the conventions of the Malaysian model of service encounters and contrast them to the model proposed by Halliday/Hasan (1985) by focusing on interactions between Chinese service providers and customers from different ethnic groups in a grocery store. A total of 120 service encounters (60 with Chinese customers, 60 with nonChinese customers) were observed and audio-taped. Analysis of the service encounters showed that Sale Request, Sale, Purchase and Goods Handover are obligatory stages in the shop, but a low frequency of Greeting, Sale Initiation and Finis was found. The infrequent use of politeness features resembling interpersonal interactions indicates a taskfocussed interaction between the service provider and customer. The results also indicate some in/outgroup differences in the Chinese service providers’ interaction with their Chinese and non-Chinese customers. The service providers were more likely to engage in Greeting and Finis with Chinese customers, and Sale Initiation with non-Chinese customers, indicating a clearer service provider-customer role and a stronger task-focus in interactions with outgroup members. The stages that are more likely to be enacted non-linguistically are Sale Request, Purchase and Goods Handover. In the grocery store, customers often brought goods they wanted to purchase to the counter, making verbalisation of Sale Request unnecessary. The non-Chinese customers were more inclined to make non-verbal Sale Requests than Chinese customers, whereas service encounters with Chinese customers had relatively more frequent non-verbalised Sale Compliance and Sale stages. The possible relevance of in/outgroup relationship on non-verbal communication in service encounters needs further investigation.


Introduction
Service encounters between customers and service providers involve several essential stages for the exchange of goods and services (Halliday/Hasan 1985;Mitchell 1957;Ventola 1987).Ventola (1987) highlighted that the non-linguistic realisation of the goods handover stage based on her study of service encounters in Finland.In contrast, Halliday and Hasan's (1985) schematic structure of service encounters was formulated based on their observations of service encounters in Australia.The handing over of goods is accompanied by verbal expressions such as "here you are".The difference is probably due to Finland being a high-context culture where meaning resides in the physical context and Australia being a low-context culture where most of the meanings are verbalised (Hall 1976).Although other studies have indicated that non-verbal communication occurs in both high-and low-context cultures (e.g.McKechnie et al. 2007 in Dubai; Moore 2008 in the United States), they have not paid attention to the linguistic and non-linguistic enactment of stages in service encounters.
Other researchers who have focused on the staging of service encounters have primarily studied the omission of opening and closing stages in interactions between Asian customers and service providers (e.g., Bailey 2001;Kong 1998;Kuang et al. 2011;Liu 2009;Pan 2000;Ting et al. 2012), and interpreted the omissions as lack of politeness in instrumental exchanges.This interpretation is reasonable given that other researchers have found elaborate sequences for politeness in other settings.The fi ndings show that service encounters in both high-context cultures such as France and Syria (Traverso 2006) and low-context cultures such as Ireland (Binchy 2005) and Uruguay (Márquez Reiter 2006) have elaborate politeness sequences.However, what distinguishes Asian countries from countries such as France, Syria, Ireland and the United States is power distance (Hofstede 1983).Hofstede conceptualised Asian countries as high power distance cultures where there are wider status gaps and more inequality in relationships.In other words, the status difference between service providers and customers is greater in high power distance Asian cultures compared to Western cultures.Putting together context-dependency in communication style (Hall 1976) and power distance (Hofstede 1983), Asian countries are high-context dependent and high power distance cultures as opposed to countries like the United States, the United Kingdom and Australia which are clearly cultures with low-context dependency and low power distance.Norms and conventions of service encounters in high-context and high power distance cultures have yet to be fully described.Thus far, Hofstede's typology of power distance has mostly been applied in business studies (Kirkman et al. 2006), but not in research on language use.In this paper we show that Hofstede's (1997) typology of national cultures can explain some differences in conventions of service encounters in Asian cultures pertaining to the linguistic enactment of the stages of the service encounter and how the status difference between service providers and customers is refl ected in the interaction.
The study aims to describe the conventions of the Malaysian model of service encounters and contrast them to the model proposed by Halliday/Hasan by focusing on grocery store interactions of Chinese service providers with their Chinese and non-Chinese customers.

Service encounters
Service encounters involving the exchange of goods and services between customers and service providers is one of the routine transactions in society so much so that service encounter is a genre of its own with universal characteristics in the staging and language features (McCarthy 1998(McCarthy , 2000;;McCarthy/Carter 1994).The pioneering study by Mitchell (1957) on three types of service encounter in Cyrenaica and Libya showed that auctions differ greatly from market and shop transactions.Auctions have four stages (auctioneer's opening, investigation of object for sale, bidding, conclusion), whereas market and shop transactions have fi ve stages (salutation, enquiry as to object of sale, investigation of object of sale, bargaining, and conclusion) (Mitchell 1957: 176, 178).Subsequently, Halliday/Hasan (1985) described green grocer interactions in Australia to be proceeding in this order: Sale Request, Sale Compliance, Sale, Purchase and Purchase Closure.The customer requests the goods (services) and the service provider responds with positive Sale Compliance if the goods are available or negative Sale Compliance if the goods are not available.In the case of positive Sale Compliance, the service provider shows the goods and names the price of the goods, and the customer pays for the goods.The service encounter closes with an expression of thanks by the service provider.These are the obligatory stages of a service encounter, but there are other optional stages such as Greeting, Sale Initiation where the service provider shows readiness to serve the customer, Sale Enquiry where the customer requests more information on the goods, and Finis where the service provider makes courteous comments to establish a continuous interpersonal relationship after the goods exchange has been completed (Halliday/Hasan 1985).Ventola (1987) challenged the fi xed sequencing of the stages based on her research in Finland on 12 service encounters in a post offi ce, a travel agency and a small shop between customers and service providers who are not acquainted.Ventola also emphasised the recursivity of Sale Request, Sale Enquiry and Sale Compliance and used fl ow charts to represent stages of service encounters.In addition, Ventola stated that some stages are enacted nonverbally, particularly the handing over of goods.Halliday/Hasan (1985) did not include goods handover as a stage in their schematic structure of service encounters because their framework consists of explicitly verbalised stages.The terms used by Ventola to refer to the stages are slightly different, and the equivalent terms used by Halliday/Hasan are placed in square brackets here: "Greeting, Turn-allocation [Sale Initiation], Service bid [Sale Compliance], Service [Sale Request], Resolution, Pay [Purchase], Goods Handover, Closing [Purchase Closure] and Goodbye [Finis]".Resolution refers to the decision of whether or not to buy the product.Martin (1985) reconceptualises the service encounter as sub-genres differentiated on the grounds of whether it has an appointment or not, and the unappointed service encounters are further differentiated in the items purchased (goods/ services) and placement of items (on display/across counter).The canonical patterns of linguistic realisation of service encounters make it useful for training of business processes and services in work (Clarke/Nilsson 2008) and investigation of other service encounters (e.g., Togher et al. 1997).However, the staging of the service encounters can vary depending on context.

High-and low-context cultures
Cultural context can be seen as low-and high-context, following Hall's (1976) typology.Hall gives examples of low-context cultures as American, German and Swiss and high-context cultures as American Indian, Arabs, Chinese, Japanese and Spanish Mexicans in New Mexico.In a low-context culture, most of the meanings are in the explicit code, and speakers would get to the point, and do not expect the listener to deduce meanings from contextual cues.In contrast, "a high-context communication or message is one in which most of the information is either in the physical context or internalized in the person, while very little is in the coded, explicit transmitted part of the message" (Hall 1976: 79).
"High context communication depends on sensitivity to nonverbal behaviors and environmental cues to decipher meaning, while low-context exchanges are more verbally explicit, with little reliance on the unstated or nuanced" (Roter et al. 2006: 28).Sanchez-Burks et al. (2003) found that Koreans and Chinese are good at interpreting indirectness in both work and non-work situations but not Americans.High-context transactions are "economical, fast, effi cient, and satisfying", but this is usually possible after individuals have been programmed to each other over time (Hall, 1976: 88).The meanings supplied by the context make "simple messages with deep meaning fl ow freely" in high-context cultures (Hall 1976: 98).The amount of contexting may infl uence reliance on words for the enactment of different stages of service encounters, and it is possible that there is more non-verbal communication in high-context cultures.
Hall observes that there are greater distinctions between insiders and outsiders in high-context cultures.He also maintains that low context cultures are characterised by "little involvement with the people" (Hall, 1976: 35).Hall (1976: 14) furthermore stresses that "words used in negotiation are not as important as negotiator status" in high-context cultures.In describing her experience of teaching business communication in Russia, Hagen (1998: 123) points out that "many highcontext cultures relegate foreigners to outsider status for many years if not forever."In service encounters, customers and service providers relate to each other in an intergroup relationship as opposed to an interpersonal relationship.
Given the background on high-and low-context cultures, we return to the foundational studies on service encounters.Mitchell's (1957) study was in Libya, a high context culture (Abubaker 2007;Hofstede 1997).Ventola's (1987) study was conducted in Finland, also a high context culture (Nishimura/Nevgi/Tella 2008), which may explain why Ventola highlighted the non-verbal goods handover stage.However, Australia, where Halliday/Hasan's (1985) observations of service encounters were conducted, is a low context culture (Poon/Evangelista/Albaum 2005; Singelis/Brown 1995).Therefore, it is not surprising that the generic structure potential framework given by Halliday/Hasan (1985) shows linguistic enactment of all possible stages in a service encounter.
Recent studies have shown that non-verbal communication is present in some aspects of service encounters in both low-and high-context cultures.Moore's (2008) study in California, a low context culture, showed that customers use gesture as a complement, supplement, and alternative to verbal communication to place their orders at a quick print shop.The canonical way of plac-ing orders at the shop is naming the document service wanted, but if there are troubles in recognition, the customers may use iconic gestures with verbal descriptors.In Moore's study, references take the form of deictic gesture (e.g., placing hands on stacks of documents, pointing) or deictic term (e.g., this) to refer directly to the service wanted.Non-verbal communication is also present in service encounters in high-context cultures (e.g., Kuang et al. 2011;Raevaara 2011;McKechnie et al. 2007).Kuang et al. (2011) analysed the openings and closings of service encounters at six Malaysian government hospitals.Kuang et al. reported that the hospital staff used non-verbal communication for nine out of 68 openings and nine out of 46 closings -mainly through smile, handshake and head nodding -and these were among the interactions categorised as polite.Interaction in partly self-service shops like convenience stores in Finland, a high-context culture, is also silent to a high degree (Raevaara 2011).McKechnie et al. (2007) used mystery shoppers to observe service providers' behavioural actions accompanying listening at 50 retail outlets at a Dubai shopping mall selling high and medium priced merchandise, and found that "making eye contact" occurred more frequently than "head-nodding" or "leaning forward/backwards".McKechnie et al.'s analysis showed that medium-priced outlet providers are more intent on listening to customers, and service providers interrupted female customers more.These nonverbal behaviours are used as indicators of liking in other studies (Palmer/Simmons 1995).In other fi elds, nonverbal communication has been largely studied in terms of its impact on customer satisfaction and service quality (e.g., Bittner et al. 1990;Gabbott/Hogg 2000;Mattila 2000;Yuksel 2004).
Research on the impact of cultural context on service encounters has revealed cultural variability in notions of politeness.Traverso's (2006) study on service encounters in two high-context cultures, France and Syria (Hall/Hall 1990;Al-Olayan/Karande 2000), showed that politeness is achieved through different elements in the interaction.The shops were those selling nonfoodstuffs and non-everyday consumer products.Traverso stated that the French show politeness through many linguistic devices such as minimisation ("That's nothing" in response to customer's "Thank you"), indirect formulation of sale request and softening devices, particularly use of past and conditional tense and mood ("I would have liked").Politeness in Syrian interactions in Traverso's (2006: 119) study took the form of a longer opening welcoming sequence that refl ects closeness and familiarity, and an acknowledgement of acceptance of the product which marks the end of the transaction usually in the form of a wish that "the new article may serve [the customer] well and happily".Traverso (2006: 118) cited Kerbrat-Orecchioni (2001) to explain that "in French shops, objects are not bought or sold, they are 'taken' or 'given': this is a way of 'euphemizing' the commercial aspect of the relationship, obscuring its 'venal' quality", but the Syrian service providers and customers also avoided the verb "to buy".Interactional closeness is also evident in service encounters at a company offering carer service in Montevideo, Uruguay (Márquez Reiter, 2006).Based on her study of 42 openings of telephone calls to the carer service company, Márquez Reiter reported that these Spanish calls were different from English calls because they were more greetings and self-identifi cations.Márquez Reiter explained that these are signs of politeness to show interest in the addressee in Spanish.Researchers have not identifi ed the Uruguayan communication style but Spain is in the middle on the high and low-context continuum (Kim/Pan/Park 1998).Márquez Reiter's fi ndings on the customers and service providers relating as acquaintance or friends in terms of self-disclosing strategies concurred with Márquez Reiter and Placencia's (2004) fi ndings on Montevidean face-to-face service encounters in shops selling clothes and household appliances.A study in Ireland, a low-context culture, shows that politeness is signaled through the use of "please" rather than politeness features that resemble interpersonal elements like in high-context cultures (Binchy 2005).From the analysis of 131 service encounters in Limerick, Binchy found that use of "please" is usual with unacquainted customers, but unusual with acquainted customers, suggesting that relationship affects the use of politeness markers.Martin/Adams' (1999) fi ndings rule out race as a variable infl uencing service providers' likelihood to thank, but this was based on data collected from 223 dyadic service encounters in shopping malls in Atlanta, Minneapolis and Wichita.Instead gender and age signifi cantly in-fl uenced thanking behaviour in that service providers tended to thank female customers and older customers, and female service providers were more likely to exhibit thanking behaviour than male service providers.Martin and Adams also cited the fi ndings of Martin (1990) and Martin and Smart (1994), who reported that service personnel failed to thank customers in 34% and 14% of face-to-face and toll-free corporate hotlines respectively.
However, studies on other high-context cultures involving customers and service providers of Asian origin have found minimal staging of service encounters that omits stages not essential for the business to be transacted, particularly openings and closings.For instance, Kuang et al. (2011) analysed openings and closings of service encounters at six Malaysian government hospitals and showed that the Malay staff initiated openings in 68 (46.57%) out of 146 transactions observed (the remainder were initiated by the patients) and thanked patients in only 46 transactions (31.51%).Kuang et al. considers "okay" as a semi-polite closing compared to "thank you", which is categorised as a polite closing.Interactions categorised as impolite were those that focused on the task only (54.41% of openings and 15.22% of closings).Another study (Ting et al. 2012) examined 422 service encounters involving Chinese service providers in four retail settings (grocery shop, photograph developing shop, bakery and fruit stall) in Sarawak, Malaysia.The study showed that the presence of verbalised openings in 39.57% of service encounters and closings in 28.44% of the encounters.The openings included greetings, sales initiation, and personal conversation whereas the closings were thanking expressions.In addition, there was evidence of in/outgroup difference in greeting and thanking behaviour because the Chinese service providers in the study were more inclined to greet and thank non-Chinese customers than Chinese customers.Explicit verbal greetings or closings also do not occur in every service encounter in Bailey's (2001) study on service encounters involving African American and Korean customers in a convenience store operated by immigrant Koreans in California.Bailey reported that the interactions with the 13 Korean customers were shorter whereas the interactions with the 12 African American customers had interpersonal involvement.Service encounters are also characterised by the absence of greeting and farewell in government offi ces and banks in Guangzhou, China (Pan 2000) and in a range of retail settings in Hong Kong (Kong 1998;Liu 2009).
The omission of openings and closings involving service providers and customers of Asian descent results in a minimal staging of service encounters and the appearance of impoliteness when compared with service encounters infused with linguistic devices to show politeness and elaborate sequences in greetings, self-identifi cations and closings in other high-context cultures.Liu (2009) attributed the lack of facework to the need for effi ciency in commercial transactions and the outsider relationship between customers and service providers whose common goal is to transact business and not to build interpersonal relationships.For example, the customers and service providers in a small shopping centre and the local market in China talked about the price in most of the 38 service encounters in Orr's (2007) study.Orr (2007: 98) reasoned that since customers and service providers interact as in/outgroup members, "a relationship between outgroup persons is only possible when the individual goals of the participants as customer and service provider have been satisfi ed in a transaction", but there was no evidence of interpersonal elements in the service encounters.Hall (1976) observed that there are clear distinctions between insiders and outsiders in high-context cultures, and customers and service providers are considered outgroup members, making interpersonal elements such as politeness features superfl uous.Hall's (1976) typology of context dependency alone cannot account for the presence of politeness features because Asian service encounters are different from those in other high-context cultures, and needs to be juxtaposed with the power distance dimension of Hofstede's (1983) typology of national cultures.

Power distance
In Hofstede's (1983) typology of national cultures, Asian countries are clearly high power distance cultures whereas countries in the West are at the lower end of the continuum.Hofstede (2003) describes small power distance countries such as Austria, New Zealand, Finland, Great Britain, and USA as having interdependent relationships between bosses and subordinates where there is more status equality.High power distance countries like Malaysia, Arab countries, Indonesia and France have dependence relationships where there are wide status differences.
The status difference between customers and service providers is clear in high power distance Asian cultures where customers expect to be served.A case in point is Mattila's (2000) study on 75 service encounters in the front offi ce and fi ne dining restaurants in a fi rst class Singapore hotel.The Asian customers gave lower quality ratings than Western customers; Asian customers focused on the process whereas Western customers focused on the outcome.Mattila explained that Asian customers expect service providers to give them personal attention and customise the service.In this sense, Simintiras/Thomas' (1998) statement that the customer is like a king and the service provider is like a beggar in some contexts is probably more apt for Asian cultures.

Method
In this section, the research site is described along with the data collection and analysis procedures.

Research site
The recordings of naturally occurring interactions were conducted in a grocery shop operated by Chinese service providers in Miri City, located in the northern region of the Malaysian state of Sarawak.Sarawak is located on Borneo Island, bordered by Brunei Darussalam in the north and Kalimantan Indonesia in the east (Figure 1).Sarawak's population of 2.47 million consists of 27 ethnic groups, the largest being the Iban (28.87%), followed by the Chinese (23.38%) and Malay (22.99%) (Department of Statistics Malaysia, Sarawak 2012).Besides the Iban, there are numerous indigenous peoples of Sarawak such as Bidayuh, Melanau, Penan, Kelabit, Kayan, Lun Bawang and Punan, each with their distinctive languages.The Chinese are divided into subgroups such as the Foochow, Hakka, Hokkien, Teochew, Cantonese, Heng Hua and Hainan, with their respective languages but they share the standard Chinese language, referred to as Mandarin.The ethnic diversity in Sarawak makes language choice a conscious decision in service encounters involving unacquainted customers and service providers (Lau & Ting 2013).The grocery store was situated in a residential area with a high density of Chinese and indigenous residents.The store sold stationery, toiletries, cleaning agents, pesticides, dried goods such as canned food, processed food, beverages, instant food and drinks, and perishables such as eggs, milk, bread and buns.A Chinese couple in their late fi fties operated the shop, assisted by their daughter in her early twenties.The couple was in charge of the management of the shop, and all three of them served customers at the counter and did the following daily tasks: stacking and arranging goods, packaging and displaying goods, ensuring cleanliness of the shop, and observing customers to prevent shoplifting.When the shop opened at 6 a.m., they pulled out the ice-cream machine, brooms and dustpans and other display racks containing onions, coals, fertilisers and garlic to display outside the shop.Before closing the shop at 7 p.m., they had to push all these back into the shop.
The Chinese couple and their daughter could speak English, Bazaar Malay, Iban, Mandarin and languages of some Chinese sub-groups, namely, Foochow, Hakka, Hokkien and Cantonese.The customers included Chinese, Malay, Iban and other indigenous peoples of Sarawak.
The shop was busy throughout the week.Every day, customers would buy some groceries after purchasing vegetables and meat at the nearby store or after taking their breakfast and lunch at the neighbouring coffee shops.Young people who attended after school classes and ping-pong training sessions also dropped into the grocery store.The other shops in the area include a drug store, a grocery store, hair saloons, a furniture shop, a car air-conditioner repair shop, a Buddhist association, a church and other offi ces.The shop usually had fewer customers on Sunday afternoons because there were fewer activities in the area.

Data collection procedures
Audio recordings of interactions in the grocery store spanned four days (2-5 February 2012) from 6.30 a.m. to 6.30 p.m. as the shop opened from 6 a.m. to 7 p.m.The data comprised 120 service encounters: 60 with Chinese customers and 60 with non-Chinese customers.The ethnic group of the customers was identifi ed based on their appearance and the language used.In the Sarawak context, physical appearance is often used to gauge ethnicity and it often becomes a basis for language choice in service encounters (Ting/Chong 2008).The service encounters with Chinese customers were in Mandarin whereas those with non-Chinese customers were in Malay (55 interactions) and English (fi ve interactions).The number of encounters for the two groups of customers was equalised to ease comparison of frequencies.There were no attempts to select particular types of service encounters; the data collection proceeded until an equal number was reached.
An MP3 player was used to record the service encounters and the fi rst researcher was present as an observer.Observation notes were taken of the nonverbal communication during the service encounters.For example, a customer pointing to a particular brand of cigarettes behind the counter or when a customer picked up the item they wanted to buy and brought it to the counter.Other gestures such as stretching of arm to show directions to items, using their fi ngers to indicate numbers and walking to display to retrieve requested items were also noted.A sign "Recording in Progress for Research" was displayed at the counter to inform customers, but they were not explicitly told of the recording in order not to disrupt the natural communication.Furthermore, since the interaction is short, it is not possible "to let customers 'get used to' the fact that they are being recorded" (Mattsson/den Haring 1998: 420).

Data analysis procedures
The audio recordings of service encounters were transcribed using the coding system of Eggins and Slade (1997).Some of the main conventions used were square brackets [ ] for inserting comments on non-verbal behaviours observed, brackets { } for translated utterances, and parentheses ( ) for denoting untranscribable talk and words placed within the parentheses were the transcriber's guess.Utterances in languages other than English were italicised in the transcripts.For the transcripts, service encounters in Mandarin were romanised (e.g., Mei you huo ah?).The English translation was put within brackets -{No stock ah?} -on a subsequent line.Transcription for service encounters in Malay was not a problem because the language uses the Roman alphabet.The data for the study amounted to 420 minutes and 4408 words.The transcripts were indexed from 1 to 120 (e.g.Transaction 1) for ease of reference.
The transcripts were analysed using the Halliday/Hasan's (1985) framework of service encounters (Appendix 1).To this framework, Ventola's (1987) goods handover stage was added to take account of the non-linguistic realisation of this distinct stage of service encounters.Crosschecking between the two researchers was done to ensure correct identifi cation of stages of the service encounter.
Three stages posing initial confusion were Sale Compliance, Sale Enquiry and Sale Request.Based on the defi nition of Sale Compliance, this is the action of the service provider to grant or reject the customer's sale request but Halliday/Hasan's (1985) examples of the service provider inviting the customer to make more purchases ("Yes, anything else?" or "Will that be all?") were not found in the dataset.Instead, Sale Compliance was indicated verbally by utterances such as "got, over there" (Transaction 96) or non-verbally by retrieving the items from the display shelves.Later we used the context to read the purpose of the service provider's response.As long as the service provider indicated that he/she was willing to sell the item to the customer and the item was available, this was considered positive Sale Compliance.Negative Sale Compliance was easy to code because it often took the form of "not available" or "don't have".
Initially it was diffi cult to distinguish between Sale Request and Sale Enquiry.Sale Enquiries are concerned with some attributes of goods contemplated for purchase and includes utterances by both the customer and service provider.The customer's sale request sometimes took the form of a question that looked like a Sale Enquiry.For example, when the customer asked "how much for ten?" (Transaction 83), it was not a sale enquiry, but a sale request because he promptly paid for the item after the price was named.In other service encounters, a question on the price was functionally a Sale Enquiry because a customer only fi nally decided to buy the item a few turns later after consulting his wife (Transaction 80), whereupon he confi rmed the initial Sale Request by putting it on the counter.In high-context cultures such as Malaysia where some stages in service encounters may not be verbalised, coding relies a great deal on the context which supplies the meaning not expressed in words.With repeated readings, the meanings and the functions of the stages usually became clearer.
The presence or absence of stages in the service encounters was analysed and the frequencies tabulated.The results were compiled separately for service encounters with Chinese and non-Chinese customers to fi nd out if there is a difference in staging for ingroup and outgroup communication, and particularly for the role of nonverbal communication.In the Results and Discussion section, excerpts of the service encounters are included to illustrate the customer-service provider interactions.

Results and Discussion
This section presents the typical structure of grocery store service encounters and compares the frequency of stages in interactions with Chinese and non-Chinese customers before focusing on the non-linguistic realisation of stages in the service encounters.

Typical grocery store service encounters
Most of the service encounters in the shop were handled by the Chinese couple who owned the shop.The male owner (Service Provider 1) handled 75 transactions, his wife (Service Provider 2) handled 41 transactions and the female sales assistant (Service Provider 3) handled four transactions.An examination of the transcripts showed that the three different service providers do not interact differently with the customers as far as the staging of the service encounter is concerned.
A typical service encounter in the grocery store took place in this manner: a customer walks into the shop and makes a Sale Request; the service provider gets the goods from the display shelves (Sale Compliance); the customer asks for the price of the goods; the service provider states the price or shows it on the cash register (Sale); the customer pays for the goods (Purchase), and the service provider hands over the goods to the customer (Goods Handover).Excerpt 1 illustrates such a service encounter.In Transaction 72 between the male shop owner (Service provider 1) and a male Malay customer, there was no explicit Greeting or Sale Initiation by the service provider because the customer walked into the shop and initiated the transaction by asking whether the shop sold envelopes.What is not obvious from the transcript is the male shop owner making eye contact with the customer, a usual way for service providers to signal engagement with the customers rather than nodding to each other to initiate contact.In Turn 5, after the service provider was sure that the customer wanted to buy an envelope (Sale Request), he took an envelope from the shelf and showed it to the customer who said "Ah", equivalent of "Yeah".The customer took the envelope and put it on the counter showing that he wanted to buy it -his action completed the Sale Compliance sequence.In other service encounters, the customer completes the Sale Compliance by making sounds such as "emm" and "ah" to confi rm that the item shown is the one he/she wanted to buy.In Excerpt 1, the customer took another item he wanted to buy from the display shelves and put it on the counter (Turn 7).This constitutes a non-verbalised Sale Request and the service provider's willingness to sell the item to him is assumed.The service provider proceeded to state the price (Sale), after which the customer paid for it (Purchase).In the fi nal turn of the service encounter, the service provider put the items in a plastic bag, handed it to the customer who took it and left the shop.This is the goods handover stage described by Ventola (1987).
Note that when the customer asked "got envelope?" in Turn 2, the service provider repeated the word "envelope?" in a questioning tone to ensure that he had heard the customer correctly (Turn 3).Hewitt (2002) studied repetition in the payment sequence of 12 service encounters at a garage and a veterinary clinic, and cited Tannen (1989: 52), who stated that repetition "bonds participants to the discourse and each other" and "provides a recourse to keep talk going where talk itself is a show of involvement, of willingness to interact, to serve positive face."Kuroshima (2010) highlighted repetition in service encounters.Kuroshima analysed 25 ordering events involving nonnative Japanese customers in a sushi restaurant, and reported that the chef repeated almost the same words to acknowledge that he had heard the order in the request-acceptance sequence.In the present study, repetition serves functions similar to the request-acceptance in Kuroshima's study in ensuring that the order for the product or service is accurately noted for further action.

Staging of service encounters in grocery store
Table 1 shows the frequency of stages in service encounters between Chinese service providers and Chinese and non-Chinese customers, tabulated based on the presence of these stages in each transaction.The frequency does not refer to the number of times a particular stage occurred, meaning that even if there are two Sale Requests in Transaction 120, it is tabulated as this service encounter having a Sale Request stage.
Out of 120 service encounters analysed, 92 were successful business transactions, indicated by Goods Handover (Table 1).Necessarily, there should also be 92 customers paying for the items (Purchase) but the frequency is only 91 because Transaction 87 involved the customer coming back to exchange a wrong item for another item of the same price.Hence, no payment was involved.Since 120 service encounters were observed, there should be 120 customers requesting items they wanted to buy, but only 117 Sale Requests were identifi ed.This is because in three of the service encounters (Transactions 9,36 and 71), customers walked into the grocery store, greeted the service provider, browsed for the items, and left with further contact with the service providers.Other than these three stages (Goods Handover, Purchase, and Sale Request), the other stages of the service encounter were lower in frequency, indicating that these can be considered the obligatory stages of service encounters.In a minimal service encounter, customers state the items they wish to buy, pay for them, and take away the goods they paid for.In Halliday/Hasan's (1985) framework of service encounters, the service provider must respond to the customer's Sale Request by granting it (positive Sale Compliance) or rejecting it (negative Sale Compliance).Altogether 117 Sale Requests were identifi ed, but there were only 60 positive Sale Compliance and 32 negative Sale Compliance (Table 1).These results should be read together with the number of service encounters where the customer bought some items (92) and the number of interactions where the customer left without buying anything (28) -referred to as successful and unsuccessful service encounters respectively.Altogether there were 92 successful service encounters, but only 60 positive Sale Compliance because the service provider's Sale Compliance is assumed in service encounters where the customers had already picked the items they wanted to buy and placed them on the counter, in which case the next step is for the customers to settle the payment for the goods.There is no reason why the service provider should not want to sell the goods to the customer.This is an example of how the context supplies meaning, making the use of words redundant.On the contrary, although there were 28 unsuccessful service encounters, the results showed 32 negative Sale Compliance.The numbers are not exactly the same because in some service encounters like Transaction 57, the customers requested more than one type of goods, giving rise to the presence of both positive and negative Sale Compliance in the same interaction.However, in four other transactions (Transactions 9,36,38,48) the custo-mers browsed for goods, but left without buying anything or interacting further with the service provider.Hence, there is no record of the service provider rejecting the customer's request for certain goods.Negative Sale Compliance is usually accompanied by suggestions of alternatives.For example, in Transaction 4, the customer wanted to buy cough lozenges and brought a packet to the counter and asked for the price.The service provider said that the lozenges did not have a good taste, after which the customer asked for another type with a white packaging.The service provider said that she did not have that particular type and suggested one with a good taste.The customer did not like that.The service provider promoted another type of good-tasting lozenges and the customer agreed to buy it.Sometimes the customer reject all the alternatives offered, in which case the service provider may advise them to try another shop.In Raevaara's (2011) study on interactions in a Finnish convenience store, the sales clerk either advised the customers to go to another shop or offered an overt explanation for the unavailability of the goods, but the latter seldom occurred in this study.
Another point of difference in comparison with Halliday/Hasan's (1985) framework of service encounters is the obligatory verbalisation of the cost of the goods by the service provider (Sale).Table 1 shows 66 service encounters with the Sale stage, yet there were 92 successful service encounters, showing that in about one-third of the transactions, the cost of the goods is not explicitly stated.Four reasons account for this.First, the goods had price labels on them.Second, the cash register showed the total amount the customer needed to pay.Third, goods like cigarettes were sold at controlled prices, and the customers already knew the prices.Fourth, the customer requested the goods according to the price (e.g., "Ada Celcom lima ringgit kah?" meaning "Do you have a Celcom reload card worth fi ve ringgit?").All these made it redundant for the service provider to explicitly state the cost of the goods.Occasionally the customer prompted the service provider's Sale statement by asking "How much is it?"The price was always stated without the use of "please" as a politeness marker, unlike Binchy's (2005) fi ndings on Irish English service encounters in Limerick, Ireland.As mentioned above, Binchy found that it is usual for service providers to use "please" with unacquainted customers, but it is not necessary for acquainted customers because the interpersonal relationship takes precedence over the business at hand.
The results show low frequencies for stages of the service encounter that resemble interpersonal elements in the opening (Greeting, Sale Initiation) and closing (Purchase Closure, Finis).Greetings in this setting usually took the form of "Hi lau bang" or "Hi lau bang nian", equivalent to "Hi, [male] boss" and "Hi, [female] boss" respectively.Examples of Sale Initiation are "Ni yao she me?" in Mandarin and "apa lu mauk?" in colloquial Malay, literally translated as "What do you want?" Greetings and Sale Initiation are infrequent because customers walk into the shop knowing what they want to buy, and may indicate this by making a verbal request or putting the goods on the counter.In other settings like Uruguay (Márquez Reiter 2006), greetings in service encounters is the norm.The results showed that explicit greetings were more frequent in interactions involving Chinese customers whereas Sale Initiation took place more frequently in interactions involving non-Chinese customers -a plausible explanation for this will be posited in relation to results on the closing of service encounters.
As for closings, expression of gratitude in the Purchase Closure stage took place in 37.5% of the 120 service encounters.For interactions with Chinese customers conducted in Mandarin, "thank you" was usually said in Mandarin (xie xie), but occasionally in English.The service providers tended to thank non-Chinese customers in English although the interaction was in Malay.Finis, a display of good will to indicate continuity in interpersonal relation, took a different form in this study.Unlike Halliday/Hasan's (1985) framework where Finis took the form of cliché good will expressions like "Have a good day", there was only one such expression, "Happy New Year", in this study.The segments of the interaction categorised as Finis were non-task conversations on topics such as the lion dance, the purchase of shop-lots and upcoming competition which has the function of strengthening the interpersonal relationship between the service provider and customer.A lion dance is a cultural performance where young men in lion dance costumes mimic the moves of lions to the beat of drums during the Chinese New Year celebration and other auspicious events for the Chinese community such as weddings and opening of a new shop.Finis tended to take place while the service provider was counting the money.This non-verbal action provides a space for a shift to non-task conversation, based on Raevaara's (2011) study on service encounters at a Finnish convenience store.Finis was more frequent with Chinese customers, probably because they were regular customers, indicated by the topics which suggested prior shared information and the service provider's initiative at starting the conversation.The non-task conversations is a type of rapport-building behavior, referred to as "connecting" by Gremler/Gwinner (2008), and includes humour, pleasant conversation and friendly interaction.From their analysis of 388 recalls of service providers' rapport building behaviours in Midwestern United States, Gremler/ Gwinner found that connecting and uncommonly attentive behavior were the most frequently mentioned and were useful to cultivate rapport.Rapport is so important to positive sales outcomes that service providers work at the small talk, including assessing their quest to present a favourable impression of themselves, the company, and the product (Clark/Drew/Pinch 2003).Hewitt (2002: 85) cites Iacobucci (1990) to assert that "nominally relation-oriented talk is not always indicative of relational goals but can be used as a strategy to achieve a task goal." The results on the low frequency of openings and closings is similar to Kuang et al. (2011) and Ting et al. (2012) on Malaysian service encounters, and supports Gao/Ting-Toomey's (1998) conclusion that there is little conventional politeness among strangers in China.However, as Greeting and Finis in Chinese-Chinese interactions were relatively more frequent, and Sale Initiation occurred more in interactions between non-Chinese customers, this suggests that the Chinese service providers were relating to their non-Chinese customers on an intergroup basis in terms of the ethnic group, and customer-service provider roles.In this light, what Hagen (1998) had written about Russian high-context culture is relevant, that is, the relegation of foreigners to outsider status for a long time.Malaysia is a high-context culture.The Chinese service providers seemed to keep their non-Chinese customers as outsiders by not engaging in interpersonal talk with them.

Non-linguistic realisation of stages in grocery store service encounters
Excerpt 2 for Transaction 16 illustrates a service encounter which was brief and conducted almost without exchange of any words: Customer 16 knew exactly what he wanted to buy, and brought the items to the counter to pay (Turns 1-2, Sale Request).The service provider keyed in the price and the total is shown on the cash register display panel (Turn 3, Sale), and the customer paid the amount shown (Turn 4, Purchase).These stages took place non-verbally, and the silence was broken by the service provider thanking the customer, who took the items from the counter and left without saying a single word throughout the whole interaction.There was an element of self-service as Customer 16 was probably a regular customer who knew the location of the items, and brought them to the counter himself.These silent service encounters are common in low personal contact retail settings where the goods on sale are uncomplicated and the service providers are not expected to promote the goods (e.g., car salesperson) or to offer their expert advice (e.g., pharmacist) as opposed to high personal contact selling like in a hair salon (Moore/Moore/Capella 2005;Toerien/Kitzinger, 2007).
Table 2 shows the extent to which stages of service encounters at the grocery store were enacted non-linguistically.The stages that need to be performed linguistically are Greeting, Sale Initiation, negative Sale Compliance, Sale Enquiry, and Finis.Sale Initiation and Finis are, by defi nition, verbal, (Appendix 1) because the service provider needs to speak to initiate the interaction with customers or to display goodwill so that customers will come back to the shop again.Positive Sale Compliance can be enacted in action by getting the items, but negative Sale Compliance has to be verbalised.Similarly, the service provider must respond verbally to customer's enquiries about the characteristics of the goods.As for Greeting, undoubtedly, the service providers and customers can nod their heads, smile or wave to greet and to bid farewell.However, no gestures such as these caught the attention of the fi rst researcher who was present at the grocery store to observe and take note of the contextual cues during the data collection.There was cursory eye contact in some interactions, but we did not consider this as a form of greeting.Kuang et al. (2011) also did not include eye contact among the actions indicating politeness (smile, handshake, head nods) in their study on service encounters at Malaysian hospitals.On hindsight, it is possible that the service provider and customer might have nodded to each other in initiating contact some of the time, and this might have escaped the attention of the fi rst researcher.A limitation of this study is that video recordings were not used, and reliance on observation notes may reduce the reliability of the description of non-verbal actions.However, even without video recordings, it may be possible to obtain an accurate description of behavioural actions with the use of a predetermined checklist, as used by mystery shoppers in McKechnie et al.'s (2007)   In this study, most of the Sale Requests were made verbally usually along the line of "Do you have.…?" and the service providers were more likely to respond verbally, usually by telling the customer where to fi nd the goods.Some of the Sale Requests were explicit requests for the items such as "give me six tins of Tigers" while others were enacted in action such as putting the items on the counter.Table 2 shows that the non-Chinese customers made more non-verbal Sale Requests (45.76%) than the Chinese customers (29.31%).This is probably because of the focus on the business at hand in the case of the former, supported by the results on the lower frequency of interpersonal elements (Greeting and Finis).The service provider granted Chinese customer's Sale Request in action more frequently than with non-Chinese customers, but the small frequencies offer limited data for analysis of differences in in/outgroup interactions.
Since there were 92 successful service encounters where the customers bought some items from the grocery store, there should be 92 Sale, 92 Purchase, and 92 Goods Handover.However, only the Sale stage was clearly seen in 66 service encounters when the service provider explicitly stated the cost of the goods, read the price tag or looked at the cash register.These were recorded in the observation notes.It is possible that the frequency of the Sale stage is underreported because some of these actions may have been missed during the observation, but, as explained earlier, there were other reasons for the absence of the Sale stage in about one-third of the service encounters.When the customers knew the price of goods, they immediately put the money on the counter or gave it to the service provider, who proceeded to key in the amount into the cash register.Another situation is when the customers made a request for goods based on the price.These two situations call into question whether the amount shown on the cash register can be considered a non-linguistic realisation of the Sale stage.However, in this study, we did not consider it as such because the customer already knew the price, but this is a point of contention that needs to be deliberated in future studies on non-verbal communication in service encounters.
The identifi cation of non-verbal communication for the Purchase stage was easier because the action of handing over the money to pay for the goods is obvious.Out of 91 successful service encounters (one involved an exchange of goods and no payment), 84.62% of the service encounters showed non-linguistic enactment of the Purchase stage.Only a few customers stated the amount they were paying or giving the service provider.
Goods Handover is a non-verbal stage in the service encounter as initially conceptualised by Ventola (1987).Interestingly, in this study, the verbal accompaniment of Goods Handover was often in the form of the service provider asking whether the customer wanted a plastic bag to keep the goods in eight service encounters.The service providers usually put the goods into the plastic bags without asking but when there were only one or two small items, they might pose the question, particularly with an increase in awareness on the non-biodegradable nature of plastic.
A comparison with other studies showed that not verbalising the Sale Request is common in kiosk encounters (Haakana/Sorjonen 2010).In their study in Finland, the customers putting their lottery tickets on the counter clearly shows the reason for the visit.Similarly, customers in Moore's (2008) study used deictic gestures to place orders at a quick print shop in California.Whether it is high-(Finland, Malaysia) or low-context cultures (United States), non-verbalisation of Sale Request is not uncommon due to the highly conventional nature of the service encounter.

Conclusion
The study on service encounters in a Malaysian grocery store shows that the obligatory stages of service encounters are Sale Request, Sale, Purchase and Goods Handover, all of which can be enacted non-linguistically.The obligatory stages are different from Halliday/Hasan's (1985) framework that specifi ed Sale Request, Sale Compliance, Sale, Purchase and Purchase Closure as obligatory.Sale Request is enacted both linguistically and non-linguistically, but the service provider's Sale Compliance is often assumed in this study because the goal of shopkeepers is to make a successful sale, indicative of the context supplying the meanings.Sale is often not explicitly verbalised and does not appear clearly as a stage in the service encounter because information on the cost of the goods is also found in the context -price tag, cash register display, and customer's familiarity with the price of goods.Purchase and Goods Handover are largely enacted in action rather than words.These fi ndings point to the important role of nonverbal behaviours and con-textual cues to carry meaning in the grocery store service encounters, as would be expected in a high-context culture.
The results on the infrequent openings and closings are expected, and concur with studies on interactions involving service providers and customers of Asian descent (Bailey 2001;Kong 1998;Kuang et al. 2011;Liu 2009;Pan 2000;Ting et al. 2012).However, what this study revealed is that these interpersonal elements are more likely to be present in ingroup interactions with Chinese customers.Their interactions with non-Chinese customers had a stronger task-focus because of infrequent Greeting and Finis (politeness features resembling interpersonal interactions) and relatively more frequent Sale Initiation.Since Sale Initiation is a behavior associated with the role of service providers, this emphasises the distinction between the service provider and the customer.The other distinction is the ethnic group distinction.With non-Chinese customers, the status difference is clearer because they are outgroups in both intergroup relations: ethnic group and customer-service provider.The fi nding is not unexpected given that Malaysia is a high power distance culture.
However, as one of the results contributing to this conclusion is the absence or presence of the Greeting stage which may take place non-verbally in the form of nods and smiles, and hence may be missed without a clearly defi ned non-verbal behaviour checklist, further studies on the nonverbal realisation of stages in service encounters are needed to reach more defi nitive conclusions.Furthermore, the study has also uncovered fuzziness in the defi nition of what counts as a Sale stage, if not verbalised, because of the information being present either in the physical context or internalised in the person (Hall 1976).Further investigations of the possible relevance of in/outgroup relationship on non-verbal communication in service encounters need to take account of these methodological issues.

Table 1 .
Frequency of stages in service encounters between Chinese service providers and Chinese and non-Chinese customers study of service encounters at a Dubai shopping mall.

Table 2 .
Frequency and percentage of non-linguistic stages in service encounters with Chinese and non-Chinese customers