Editor

JLLT edited by Thomas Tinnefeld
Showing posts with label 81 Shively. Show all posts
Showing posts with label 81 Shively. Show all posts
Journal of Linguistics and Language Teaching
Volume 2 (2011) Issue 1
pp. 15 - 51


Role Plays and Naturalistic Data
in Pragmatics Research:
Service Encounters during Study Abroad

Rebeca Bataller (Gettysburg (PA), USA) / Rachel Shively (Normal, (IL), USA)


Abstract (English)
The purpose of this study is to investigate the validity of role-play data in pragmatics research through a comparison of role-play and naturalistic interactions by L2 learners of Spanish in the service-encounter context. The data for this comparison come from two independent studies (Bataller 2010; Shively 2011), each of which analyzed the pragmatic development in service encounters of L2 learners studying abroad in Spain for one semester. While both studies shared the same pragmatic aspect which was investigated, they differed on the primary method of data collection employed: Bataller (2010) employed an open role play, while Shively (2011) collected naturalistic data. The results of this comparative study indicate that the role-play data resembled the naturalistic data in many respects, but that there were also differences which may have been due to the method of data collection employed.


Resumen (Español)
El objetivo de este estudio es analizar la validez del role-play como método de recolección de datos de elementos pragmáticos de la lengua. En el estudio se comparan interacciones naturales con interacciones simuladas realizadas a través de un role-play por dos grupos de estudiantes de español como segunda lengua. Los datos comparados provienen de dos estudios independientes (Bataller (2010); Shively (2011)). Ambos estudios analizan el desarrollo de la habilidad pragmática de estudiantes de español como segunda lengua que estudiaron en España durante un semestre. Aunque los dos estudios analizan el mismo aspecto pragmático, se diferencian en el método de recolección de datos utilizado. Bataler (2010) utilizó un ‘role-play’ y el Shively (2011) utilizó datos naturales, recogidos principalmente a través de grabaciones de interacciones. Los resultados de este estudio comparativo indican que los datos recogidos a través del role-play son similares a los datos naturales en muchos aspectos, pero también difieren en otros, lo cual parece indicar que el método de recolección de datos es un factor importante que puede influir en los resultados encontrados en estudios de pragmática.


1   Introduction

In the literature on data collection methods in pragmatics, the relative advantages of elicited and natural data represent a topic which has received considerable attention (cf. Beebe & Cummings 1996; Félix-Brasdefer 2003; Hartford & Bardovi-Harlig 1992; Kasper 2000; Rintell & Mitchell 1989). Three issues have been central to the debate: validity, comparability, and practicality. On the one hand, elicited data collection methods (e.g. production questionnaires, role plays) do not offer a direct measure of authentic discourse and may result in data which diverge in important ways from language produced in natural settings (e.g. Franch & Lorenzo-Dus 2008, Golato 2003). Naturally-occurring speech, on the other hand, represents a more valid measure of actual language use, but typically does not allow the researcher to exercise control over social and contextual variables. This lack of control makes it difficult to replicate the same situation and to systematically compare speech samples from different individuals or groups and at different points in time (Beebe & Cummings 1996).

Elicited data collection also has certain practical advantages over natural data collection, first and foremost, the ability to efficiently obtain large amounts of comparable data by means of questionnaires and role plays (Kasper 2000). An additional constraint of natural data is that some speech acts occur infrequently in spontaneous interactions or are realized in settings inaccessible to researchers, whereas such speech acts can easily be the object of elicited data collection methods (e.g. Billmyer & Varghese 2000, Cohen 1996, Kasper & Dahl 1991).

Of the methods available to elicit spoken speech acts, some studies point to role plays as a more valid measure of authentic language use compared to written production questionnaires (e.g. Eisenstein & Bodman 1993, Félix-Brasdefer 2003, Tran 2006). However, only a few studies have empirically analyzed this issue. In order to further examine the validity of role plays in pragmatics, the present study compares the pragmatic development of second language (L2) learners of Spanish using two different methods of data collection: open role plays and naturalistic discourse. Service encounters were the object of analysis and the participants were U.S. study abroad (SA) students in Spain. We begin by providing an overview of previous research concerning methodology in pragmatics and a description of service encounters (chap. 2), after which we describe the research design (chap. 3), present the results of the comparative methods study (chap. 4), its limitations (chap. 5), and discuss the implications for pragmatics research (chap. 6).


2   Literature Review

2.1 Data collection methods in pragmatics

A variety of data collection methods have been employed in pragmatics research. These include written production questionnaires, elicited conversations, spontaneous conversations, closed and open role plays, diaries, and field notes (e.g. Cohen 2004, Kasper and Dahl 1991, Kasper 2000). In studies which have compared two or more data collection methods, the findings have pointed to possible method effects. In consideration of the ubiquity of written production questionnaires (i.e. discourse completion tests or DCTs) in speech act research, a great deal of effort has been directed at comparing this and other methods (e.g. Beebe & Cummings 1996, Franch & Lorenzo-Dus 2008, Golato 2003, Hartford & Bardovi-Harlig 1992, Kasanga 1999, Rintell & Mitchell 1989, Rose 1994, Yuan 2001). However, a few studies have also examined the similarities and differences between role plays and natural data (Eisenstein and Bodman 1993; Félix-Brasdefer 2007; 1991; Tran 2006; Turnbull 2001).

Role plays elicit oral production data through simulated interaction between two or more interlocutors who act out imagined roles in a situation predetermined by the researcher (Crookall & Saunders 1989). Two types of role plays have been employed in pragmatics: closed and open (Kasper & Dahl 1991). Closed role plays involve one turn each by the research participant and the interlocutor, whereas open role plays allow participants to take as many turns as they deem necessary to complete the required task. Some role-play protocols explicitly instruct the participant to perform a particular speech act, while others simply provide a scenario without prescribing which speech acts must be performed (Félix-Brasdefer 2007).

It has been argued that role plays are a more valid representation of real-life spoken language than written production questionnaires (Félix-Brasdefer 2003; Tran 2006). Unlike questionnaires, role-play data capture many of the discourse features found in natural speech: intonation, pauses, turn-taking, overlap, and laughter, among others (Kasper 2000). Role-play interactions also elicit more negotiation, repetition, and supportive moves in comparison to written measures (e.g. Beebe & Cummings 1996, Félix-Brasdefer 2003, Margalef-Boada 1993, Rintell & Mitchell 1989).

Despite the potential advantages of role plays compared to written questionnaires, several studies suggest that role-play data may differ from natural data in certain ways, a finding which calls into question the validity of role plays as a measure of authentic language. Hassall (1997), for example, suggested that in both role-play and discourse completion test (DCT) request data, L2 learners used information from the description of the situations and, as a result, produced longer and more elaborate responses than they would in real-life situations. Likewise, Turnbull (2001) argued that in request refusals, role-play participants were long-winded, repeating themselves with more requests for information and confirmations than in the natural data. Turnbull’s role-play participants were also overly eager to respond to the request and interrupted their interlocutor more often compared to the naturally-occurring interactions. Conversely, in Eisenstein & Bodman’s (1993) study of expressions of gratitude, the role-play data were not as long and complex as the natural data.

Félix-Brasdefer (2007) also found differences between role plays and natural data. Spanish requests realized in natural settings evidenced greater variability in prosodic features, involved longer requesting sequences, and contained more idiomatic expressions than requests generated in open role plays. Naturally-occurring requests also displayed features that were absent from the role-play data, including elliptical requests and negation as a mitigator. Precursors (e.g. titles, names), however, were more frequent in role plays than in the natural data, a finding that Félix-Brasdefer argued may be related to the simulated nature of role plays and the way in which social interaction is initiated in an artificial context. Despite these indications of method effect, there were enough similarities between the two methods for the author to conclude that, “although role plays do not provide the researcher with authentic interactional data, the results of the present study showed that role plays represent an approximation to spoken discourse” (Félix-Brasdefer 2007: 177). The present study aims to further investigate the validity of role plays in measuring authentic discourse.

2.2  Service encounters in Peninsular Spanish

The speech activity in focus is the service encounter. A robust number of studies have examined service encounters in a variety of languages and cultures (e.g. Aston 1995, Callahan 2009, García 1993, Gavioli 1995, Hickey 2005, Kerbrat-Orecchioni 2005, Kidwell 2000, Márquez Reiter 2005 and 2006, Pinto 2002, Placencia 1998, 2004, 2005; Traverso 2001, Vélez 1987). Retail and food-related service encounters, as typically monotopical and goal-oriented interactions, tend to display a routine structure (e.g. Drew & Heritage 1992, Ide 1998, Kidwell 2000, Lamoureux 1988, Ventola 1987, Zimmerman 1992). In those cases in which a client must request a product from the service provider rather than obtain it in the store himself or herself, the following structural elements and their sequence are typically present:
-       (1) opening,
-       (2) request for product,
-       (3) provision of product and/or negotiation,
-       (4) payment requested,
-       (5) payment provided, and
-       (6) closing.

Due to local contingencies, not all of these elements may appear in every instance, but the list above outlines the tasks and the order in which participants routinely carry out those tasks in such encounters.

Although requests typically occur in service encounters, various authors (Danblon et al. 2005; Kerbrat-Orecchioni 2005; Sifianou and Antonopoulou 2005) point out that service-encounter requests may differ from requests in other contexts, given that it is the hearer (i.e. the service provider) and not the speaker (i.e. the customer) who normally benefits from the exchange. Danblon et al. (2005), for example, argued that, because the sales request is commercially beneficial to the provider, employees used more politeness markers than their customers, but that politeness strategies were intended to display friendliness, good manners, and courtesy rather than mitigate a threat to face.

Although retail service encounters in a variety of cultures appear to have similar structural features, cultural differences also exist (e.g. Aston 1995, Callahan 2009, Gavioli 1995, Lamoureux 1988, Márquez Reiter & Placencia 2004). For example, the way in which service encounters are realized in Peninsular Spanish differs from other dialects of Spanish (Placencia 2005). In a study of naturally-occurring corner-shop encounters in Madrid, Placencia (2005) described openings as typically brief and  found that they frequently consisted of the informal adjacency pair hola-hola (‘hi-hi’), although other more formal greetings such as buenos días (‘good morning’) were also attested. In some cases, the first pair part of the greeting was not followed by the second-pair part of the greeting, but rather by the request.

Requests for products occurred immediately after the opening sequence in Placencia’s (2005) data, reflecting an orientation to the earliest possible initiation of the transaction. Requests were realized almost exclusively with imperatives (e.g. dame un café (‘give me a coffee’)), simple interrogatives (e.g. ¿me pones un café? (‘you give me a coffee?’)), or elliptical forms (e.g. un cafe (‘a coffee’)). Want statements (e.g. quiero un café (‘I want a coffee’)) and conventionally-indirect forms (e.g. ¿me puedes poner un café? (‘can you give me a coffee?’)) were also present, but infrequent (Placencia 2005). The predominance of hearer-oriented verbs rather than speaker-oriented verbs and the frequent use of direct requests in Spanish, as reported by Placencia, has also been observed by other authors (e.g. Pinto 2002, Márquez Reiter 2000). In American-English service-encounter requests, the opposite appears to be true: speaker-oriented forms predominate, particularly want statements, need statements, and conventionally-indirect forms (Pinto 2002, Vélez 1987).

Internal mitigation of requests was infrequent in the Madrid corner-store service encounters. Spanish has a number of resources for internal mitigation, such as the diminutive (e.g., cafecito ‘little coffee’), the politeness formula por favor(‘please’), lexical downgrading of the main verb (e.g., regalar (‘to give a gift’)), syntactic downgrading (e.g., quería un café ‘I wanted a coffee), and hedging (e.g. unos seis de ésos (‘about six of those’)). In the few instances in which internal mitigation occurred in the Madrid corner-store encounters, Spaniards only used diminutives and the politeness formula (Placencia 2005).


3   Research Design

This study, for the purposes of comparison, brings together data collected from two independently-conducted studies concerning L2 pragmatic development in service encounters during study abroad (SA) (Bataller 2008, 2010; Shively 2011). In this section, a short summary of each individual study is provided, followed by a description of the comparative study.

3.1 Role-play study

The participants in Bataller’s (2008, 2010) study consisted of 31 English-speaking learners from the U.S. who studied abroad in Valencia, Spain, for four months. Twenty-four of the students were from a large East-Coast university and the remainder came from other parts of the U.S. Participants had completed a minimum of three semesters of college-level Spanish (or the equivalent) prior to SA. None had previously studied abroad or was a heritage speaker of Spanish. There were 13 males and 18 females, with ages ranging from 18 to 22 (average of 20). Thirty-two native speakers (NSs) of Peninsular Spanish, 12 males and 20 females, also participated in the study. The role of the Spanish native speaker group was to serve as the baseline data against to compare the learners' interlanguage pragmatic development.

An open role play with two service-encounter scenarios was used for data collection. Only one of the role-play scenarios was analyzed in the present study: requesting something to drink in a cafeteria. Before beginning the role play, participants received a card with the following description in both Spanish and English: “Estás en una cafetería cerca de la universidad porque quieres beber algo (agua, una Coca-cola, un zumo, un café…). Vas a la barra y le pides algo para beber a la camarera, a la cual no conoces de nada. (‘You are in a cafeteria close to campus because you want to get a drink (water, a Coke, some juice, coffee…). You go up to the counter and order something to drink from the person behind the counter. You do not know her.’)”

A 31-year-old Spanish female performed the role play (in the role of the bartender) with the SA students and Spanish native speaker (NS) participants. The role plays took place in an empty room in the center in which the SA students took classes. One participant at a time entered the room and performed the role play with the interlocutor. For the learners, data were collected at two different times: at the beginning (Time 1) and toward the end (Time 2) of the semester abroad. Data from the Spanish NSs were collected only once. All of the role plays were audio-recorded and transcribed.

3.2 Naturalistic Study

Shively’s (2011) participants included seven students (2 male, 5 female) from a large public university in the Midwest U.S., who spent four months studying in Toledo, Spain. The participants’ profile was similar to that of the students in Bataller’s study: all seven students had received a minimum of three semesters of college-level Spanish prior to studying abroad, their ages were from 20 to 21, and none of the students had previously studied abroad or were heritage speakers of Spanish.

The data consist of naturalistic audio-recordings which participants made of themselves in situ while participating in service encounters in Toledo. Students recorded at least five encounters each week during weeks two (Time 1), six (Time 2), and eleven (Time 3) of the semester, resulting in a minimum of 15 recorded service encounters per student. The full corpus includes a total of 113 service-encounter interactions.

To make the recordings, participants were instructed to carry a digital audio-recorder in a purse, backpack, or pocket, to begin recording each service encounter before entering the establishment, and to stop recording after exiting. In order to maintain the natural quality of the data, students were directed to record service encounters which they would have participated in regardless of the research project. Students made recordings in a number of different establishments in Toledo including a shoe store, a jewelry store, and various bars and cafés. Students did not have personal relationships with any of the service providers recorded. All of the audio-recordings were transcribed and analyzed by the researchers. In addition to the audio-recordings, students wrote journals each week, describing their learning of pragmatics. Journal data were analyzed qualitatively, as a means to gain insight into students’ learning.

3.3 Comparative Methods Study

In order to compare the two data-collection methods, the researchers analyzed openings, requests, negotiations, closings, and the length of the interactions. In the case of openings, requests, and closings, the 62 role plays in which students ordered a drink were compared to 59 service encounters selected from the naturalistic corpus. In the 59 naturalistic interactions which were chosen, students requested a specific product or service in cafés, bars, retail establishments, or information desks. Naturalistic interactions which did not contain requests for a specific product or service, such as interactions involving requests for information (e.g., Where is the bus stop?) and those lacking a verbal request were excluded. Although only 18 of the 59 selected naturalistic interactions were food-related, the openings, requests, and closings produced in non-food-related settings were realized in a manner similar to the naturalistic encounters in which students ordered food or drinks. For this reason, these 59 naturalistic interactions were judged as appropriate for the comparison with the role plays with regard to openings, requests, and closings.

Strategies used in openings, requests, and closings were coded and analyzed by the researchers. Request strategies were coded using a modified version of the categories originally developed by Blum-Kulka et al. (1989), shown in Table 1 below:


Strategy type
Example
Head act








Direct strategies









Mood derivable
Ponme un tinto de verano. (‘Give me a summer red wine.’)

Ellipsis
Un gofre con chocolate. (‘A waffle with chocolate.’)

Want statement
Quiero una fanta naranja. (‘I want an orange fanta.’)

Need statement
Necesito una bebida. (‘I need a beverage.’)

Simple interrogative
¿Me pones un Pepsi, por favor? (‘You give me a Pepsi, please?’)
Indirect strategies


Query permission
¿Puedo tener café con leche? (‘Can I have café latte?’)

Query ability
¿Puedes ponerme un café, por favor? (‘Can you give me a coffee, please?’)

Query possibility
¿Es posible, quiero un cubalibre, por favor? (‘Is it possible, I want a rum and coke, please?’)
Supportive moves


Lexical mitigation



Politeness marker
¿Me pones un Pepsipor favor? (‘You give me a Pepsi,please?’)


Appealer
Dame una bebida, ¿? (‘Give me a drink, yes?’)

Syntactic mitigation



Aspect
Quería una Coca Cola. (‘I wanted a Coca Cola.’)


Conditional tense
Me gustaría una Coca Cola(‘I would like a Coca Cola.’)


Mood
Quisiera un café, por favor. (‘I would like a coffee, please.’)


Formal pronoun
¿Puede usted darme una Coca Cola? (‘Can you give me a Coca Cola?’)
Verb orientation

Speaker-oriented
¿Puedo tener un café? (‘Can I have a coffee?’)

Hearer-oriented
¿Me pones agua del grifo? (‘You give me tap water?’)

Impersonal
Es posible, quiero una coca-cola (‘It’s possible, I want a Coca Cola’)
Tab. 1: Request-coding categories

In addition, the researchers compared the length of the role-play and naturalistic interactions by counting the number of words and turns per interaction. The statistical significance of differences in mean length was measured using t-tests. For the analysis of length, only 17 of the naturalistic interactions were included. Naturalistic interactions which did not involve ordering food or drinks were removed from the calculation of length due to the fact that encounters in the corpus that were realized in other settings tended to be longer than the food-related encounters and their inclusion would have biased the comparison. One particularly lengthy food-related naturalistic interaction involving small talk was also removed, leaving a total of 17 naturalistic interactions. Seventeen of the role-play interactions were then selected at random for purposes of statistical comparison with the 17 naturalistic interactions.


4   Results

4.1 Opening Phase

The opening phase consisted of the strategies used by the participants to establish contact with their interlocutor. As Tables 2 and 3 show, learners in both studies primarily employed greetings to open their interactions. However, greetings were more frequent in the role-play study (Time 1: 77%, Time 2: 100%) compared to the naturalistic study (Time 1: 74%, Time 2: 67%, Time 3: 67%). The difference in the frequency of greetings between the two studies was related to the fact that a greater number of interactions in the naturalistic data did not include an opening phase. As shown in Table 3, the percentage of naturalistic interactions with no verbal opening was 17% in Time 1, 33% in Time 2, and 33% in Time 3. In the role-play data, on the other hand, only 10% of the interactions in Time 1 were produced without an opening and all Time 2 interactions contained an opening:


Type
Time


Time 1
Time 2

No.
%
No.
%
Opening strategies




Greetings
24
77%
32
100%

hola
21
68%
24
77%

buenos días / buenas
1
3%
2
6%

hola, buenos días / hola, buenas
--
--
3
10%

hola, ¿cómo estás?
2
6%
3
10%
Alerters
4
13%
3
10%
Total opening strategies
28
90%
35
100%
No verbal opening
3
10%
--
--
Total interactions
31
100%
31
100%













Total = 62 interactions
Tab. 2: Strategy use in openings (role-play data)


Type
Week


Time 1
Time 2
Time 3

No.
%
No.
%
No.
%
Opening strategies






Greetings
17
74%
10
67%
14
67%

hola
14
61%
10
67%
13
62%

hola, ¿cómo estás?
3
13%





hola, buenas
--
--
--
--
1
5%
Alerter
--
--
--
--
--
--
Server prompt
2
9%
--
--
--
--
Total opening strategies
19
83%
10
67%
14
67%
No verbal opening
4
17%
5
33%
7
33%
Total interactions
23
100%
15
100%
21
100%











Total = 59 interactions
Tab. 3: Strategy use in openings (naturalistic data)


In the encounters which lacked an opening phase, the participants initiated the interaction by either asking a question or by directly issuing the request for a product. An example of the latter is provided in (1) below:

(1) Greta (G) at Time 2 (naturalistic study)
1       G:     um tch h (.) un gofre con chocolat y- y nata
1       G:     um tch h (.) a waffle with chocolate and- and cream

The absence of openings in some naturalistic interactions can be explained by the fact that the local conditions of real-life service encounters were variable and some situations precluded openings. For example, one student had been waiting at the bar while a busy bartender filled orders for other customers and when his turn came, he made the request directly without an opening. Where time is of the essence, going directly to the request may be acceptable or even preferred. In other cases, the presence of the customer in the establishment may have been enough to initiate the interaction or the opening may have been realized non-verbally, for example, by a nod or a smile.

Regarding the specific greeting strategies produced, the informal greeting hola (‘hi’) was the preferred greeting in both studies. This strategy was observed in 68% to 77% of the role-play interactions and in 61% to 67% of the naturalistic encounters. A notable difference, however, was the fact that alerters appeared in the role plays exclusively. Role-play participants employed alerters such as perdóname (‘excuse me’) and camarera(‘bartender’) 13% of the time  in Time 1 and 10% in Time 2. An example is provided in (2) below:

(2) Participant E7 with bartender (B) at Time 1 (role-play study)
1         E7:  buenas, perdóname, una Fanta de Naranja, por favor
1         E7:  hello, excuse me, an Orange Fanta, please

Although alerters were not observed in openings in Placencia’s (2005) naturalistic corner store interactions in Madrid, the Spanish NSs who provided the baseline data in the role-play study did, indeed, use alerters in their openings and they did so at a similar frequency as the nonnative speakers in the role-play study (see Bataller 2008).


4.2     Request Phase

4.2.1  Head Acts

Requests function to indicate the reason for the visit or to identify the business at hand. Request strategies found in the two studies are shown in Tables 4 and 5. Similarities between the two data sets include the fact that learners used the query permission puedo tener (‘can I have’) rather frequently at Time 1 (naturalistic: 33%, role-play: 35%) and notably reduced their use of this strategy by the end of SA (naturalistic: 4%, role-play: 6%). Since puedo tener is a pragmatically inappropriate request strategy, this shift represented a movement towards the NS norm (Bataller 2010; Pinto 2002; Placencia 2005). Both groups of students also increased their use of ellipsis over the course of the semester—another movement toward the NS norm—although the increase was somewhat greater in the naturalistic study. A final similarity concerns the number of strategy types: a total of seven different strategies were found in both data sets:


Strategy type
Time


Time 1
Time 2

No.
%
No.
%
Head acts
31
100%
32
101%
Direct strategies
19
61%
28
88%

Mood derivable
1
3%
2
6%

Ellipsis
1
3%
4
13%

Want statement
14
45%
22
69%

Need statement
3
10%
--
--

Simple interrogative
--
--
--
--
Indirect strategies
12
39%
4
13%

Query permission
11
35%
2
6%

Query ability
1
3%
1
3%

Query possibility
--
--
1
3%
Lexical mitigation
18
100%
17
100%

Politeness marker
17
94%
17
100%

Appealer
1
6%
--
--
Syntactic mitigation
8
102%
15
100%

Aspect
1
13%
8
53%

Conditional tense
3
38%
6
40%

Mood
3
38%
1
7%

Formal pronoun
1
13%
--
--
Verb orientation
31
100%
32
100%

Speaker-oriented
28
91%
24
75%

Hearer-oriented
2
6%
3
9%

Impersonal
--
--
1
3%

Ellipsis (no verb)
1
3%
4
13%
Total interactions
31
100%
31
100%
Total = 62 interactions
Tab. 4: Strategy use in requests (role-play data)



Strategy type
Time


Time 1
Time 2
Time 3

No.
%
No.
%
No.
%
Head acts
24
100%
17
100%
24
101%
Direct strategies
16
67%
13
76%
21
88%

Mood derivable
--
--
--
--
2
8%

Ellipsis
2
8%
4
24%
6
25%

Want statement
6
25%
5
29%
4
17%

Need statement
8
33%
2
12%
5
21%

Simple interrogative
--
--
2
12%
4
17%
Indirect strategies
8
33%
4
24%
3
13%

Query permission
8
33%
3
18%
1
4%

Query ability
--
--
1
6%
2
8%

Query possibility
--
--
--
--
--
--
Lexical mitigation
--
--
5
100%
3
100%

Politeness marker
--
--
5
100%
3
100%

Appealer
--
--
--
--
--
--
Syntactic mitigation
4
100%





Aspect
--
--
--
--
--
--

Conditional tense
--
--
--
--
--
--

Mood
4
100%
--
--
--
--

Formal pronoun
--
--
--
--
--
--
Verb orientation







Speaker-oriented
22
92%
10
59%
10
42%

Hearer-oriented
--
--
3
18%
8
33%

Impersonal
--
--
--
--
--
--

Ellipsis (no verb)
2
8%
4
24%
6
25%
Total interactions
23
100%
15
100%
21
100%
Total head acts = 65                        Total interactions = 59
Tab: 5: Strategy use in requests (naturalistic data)

Differences in strategy use were also discovered. Learners in the naturalistic study made requests less frequently with want statements (Time 1: 25%; Time 2: 29%; Time 3: 17%), while this was the single most frequent strategy used by participants in the role plays (Time 1: 45%; Time 2: 69%). The predominance of want statements in the role plays may be a consequence of the wording of the description that students read prior to performing the role play, in which the verbquerer ('to want') was used in the phrase “Quieres beber algo” (‘You want to drink something’). Students may have lifted this particular verb and incorporated it into their role-play requests. Other notable differences in strategy use include the simple interrogative and the need statement. The former appeared in the naturalistic data in Times 2 (12%) and 3 (17%), but was not present in the role-play data, and the latter was found in both studies, but was more prevalent in the naturalistic data. However, students in the naturalistic study never used need statements in food- and drink-related service encounters, only in other settings.

Finally, while both groups of students increased their use of direct strategies over time, the specific direct strategies adopted differed. Learners in the role-play study increased their use of the want statement from Time 1 to Time 2, while learners in the naturalistic study incorporated more mood derivable, elliptical, and simple interrogative requests in Times 2 and 3. Several participants in the naturalistic study indicated in their journals that they had learned about these particular pragmatically-appropriate request strategies from in-class instruction during their semester abroad.

4.2.2 Verb Orientation

Both studies display a similar development over time regarding verb orientation. That is, speaker-oriented requests became less frequent and hearer-oriented and elliptical requests became correspondingly more frequent. However, this shift was more marked in the naturalistic data, where the use of speaker-oriented requests decreased from 92% in Time 1 to 59% in Time 2 and 42% in Time 3, compared to the modest decrease in the role-play data: 91% in Time 1 to 75% in Time 2. As previous research indicates, Spanish NSs tend to use hearer-oriented strategies when making requests, whereas English NSs have a preference for speaker-oriented strategies (e-g. Blum-Kulka et al. 1989, Márquez Reiter 2000, Vélez 1987). The predominance of speaker-oriented requests in Time 1 is likely due to L1 transfer and the subsequent decrease in speaker-oriented strategies can be regarded as a movement towards the NS norm. Again, several students in the naturalistic study indicated in their journals that they had learned about pragmatically-appropriate request strategies through in-class instruction, which may account for the more substantial increase in hearer-oriented and elliptical requests in the naturalistic study.

4.2.3  Internal Mitigation

As Tables 4 and 5 indicate, there were important differences in the use of internal mitigation. The role-play students employed a greater number and wider range of mitigation devices. Whereas six different types of mitigation strategies appeared in the role plays, only two strategies - mood and the politeness marker (por favor) - were present in the naturalistic data. Syntactic mitigation was relatively frequent in the role plays at both Times 1 (26 %) and 2 (48 %), while in the naturalistic data syntactic mitigation occurred in only 17% (N = 4) of interactions at Time 1 and was not observed in Times 2 or 3. Role-play participants mitigated their requests most often with aspect and conditional tense, and usedmood to a lesser extent. Learners in the role-play study also employed the politeness marker at a substantially higher rate (in 56% of the interactions in Times 1 and 2) than learners in the naturalistic study (Time 1: 0%; Time 2: 33%; Time 3: 14%).

The more frequent use of lexical and syntactic mitigation discovered in the learner role-play data has parallels in NS data. While Placencia (2005) reported that Spaniards used internal mitigation infrequently in naturalistic service encounters, the Spanish NSs who provided the baseline data in the role-play study employed lexical and syntactic mitigation at a similar rate as learners in the role-play study (see Bataller 2008).

4.2.4  Lexical Items

A final difference related to requests concerns the use of two phrases, una bebida (‘a drink’) and algo para beber(‘something to drink’), in the role plays. Ten role-play participants made requests with these phrases, eight in Time 1 and 2 in Time 2. There are no instances in the naturalistic data of these lexical items or similarly generic terms; participants in the naturalistic study always requested specific products (e.g., a coffee, a red wine). It is likely that role-play participants lifted these particular phrases from the description of the situation that they received immediately prior to acting out the role play.

4.3  Negotiation Phase and Non-Transactional Talk

The negotiation phase followed the initial request and consisted of information exchange about the product. The negotiation phase was not always present in either the role-play or the naturalistic data. Sample (3) provides an example of negotiation from the role plays. The negotiation phase begins in line 2, following the request, when the bartender asks whether the student wants her soft drink in a can or bottle. The student responds and then the server further queries the student about ice in line 4:

(3) Participant E2 with bartender (B) at Time 2 (role-play study)
1         E2: hola! (.) gustaría una coca-cola
                 por favor   
2         B:    ¿de bote o de botella?               
3         E2: de botella, por favor
4         B:    ¿quieres hielos tambien?           
5         E2: sí, sí, hielos                    
6         E2: gracias                       
7         B:   de nada
8         E2: ¿cuánto vale?                   
9         B:   un euro
10       E2: aquí tiene             
11       B:   vale gracias hasta luego           
12       E2: hasta luego

1         E2: hi! (.) I would like a coca-cola
                 please  
2         B:    can or bottle?               
3         E2: bottle, please
4         B:  do you want ice too?          
5         E2: yes, yes, ice                  
6         E2: thanks                      
7         B:  you're welcome
8         E2: how much does it cost?                   
9         B:  a euro
10       E2: here you are                  
11       B:   OK thanks see you later         
12       E2: see you later


When negotiation occurred in the naturalistic data, it was often similar to the role plays regarding the types of information exchanged. The negotiation which occurs in (4) is similar to the role-play example in (3): the server issues requests for information in lines 4 and 6 to clarify the customer’s needs, and the student responds in lines 5 and 7:

(4) Samantha (S) with bartender (B) at Time 2 (naturalistic study)
1       S: hola
2       B: hola
3       S: ¿puedes ponerme un café (.) por
              favor?
4       B: ¿con leche?=
5       S: =con leche, sí
6       B: ¿para tomártelo aquí?
7       S: n- no, [para llevar
8       B:  [para llevar
9       S: para llevar
10      (2.3)
11      S: es que tengo mucho sueño hoy heh heh
12      B: sí::
13      S: fui a la- las fallas (.) [en Valencia
14      B: [a:y las fallas
15      B: ¿te gustaron?
16      S: sí::=
17      B: =sí?=
18      S: =pero son locos heh heh
19      B: sí: a que sí  [heh heh heh
20      S: [heh heh heh
21      B: mucho ruido, verdad?=
22      S: =sí, mucho
23      (0.5)
24      S: y fui a un- a un prime:ro-
              primera corrida (.)  [después
25      B:  [ah:::
26      S: eso también es l(hh)oco(hh)[heh heh
              heh
27      B:[sí:::
28      (7.0)
29      S: ¿cuánto cuesta?
30      B: un euro
31      S: un euro
32      (49.0) ((sounds of the espresso machine))
33      S: muy bien (.) gracias

1       S: hi
2       B: hi
3       S: can you give me a coffee (.) please?
4       B: with milk?=
5       S: =with milk, yes
6       B: to drink it here?
7       S: n- no, [to take away
8       B: [to take away
9       S: to take away
10      (2.3)
11      S: it’s that I’m really tired today heh heh
12      B: yes::
13      S: I went to th- las fallas (.) [in Valencia
14      B: [a:y las fallas
15      B: did you like them?
16      S: yes::=
17      B: =yes?=
18      S: =but they’re crazy heh heh
19      B: yes: that’s right  [heh heh heh
20      S: [heh heh heh
21      B: a lot of noise, right?=
22      S: =yes, a lot
23      (0.5)
24      S: and I went to a- a first- first bull fight (.)  [afterwards
25      B: [ah:::
26      S: that also is c(hh)razy(hh)[heh heh heh
27      B: [yes:::
28      (7.0)
29      S: how much does it cost?
30      B: one euro
31      S: one euro
32      (49.0) ((sounds of the espresso machine))
33        S: great (.) thanks

A difference, however, was that in the naturalistic data, non-transactional small talk also occurred during or after the negotiation phase in four of the 59 interactions (7%). After the negotiation in lines 4 to 9 in example (4), the student initiates a topic unrelated to the transaction in line 11: her weekend trip to Valencia. This topic is developed for 16 lines while the bartender prepares the coffee. In contrast, talk was limited to the transaction in the role plays. Since role-play participants were specifically directed by the researcher to order a drink and may not have viewed small talk as an appropriate or necessary part of that task.

4.4  Closing Phase

In the closing phase, participants wrap up the interaction with their interlocutor. Tables 6 and 7 show the closing strategies used by students in both studies. An expression of gratitude (e.g., gracias ‘thank you’) was the single most frequent closing strategy in both studies, used by itself and in combination with other strategies. Inclusive of all instances of gratitude found in role-play closings, this strategy was employed in 97% of the interactions in Time 1 and 94% in Time 2. In the naturalistic data, an expression of gratitude appeared in 78% of the interactions in Time 1, 60% in Time 2, and 57% in Time 3. As these frequencies indicate, an expression of gratitude was preferred more strongly in the role-play study (Tables 6 and 7):


Type
Time


Time 1
Time 2

No.
%
No.
%
Closing strategies





Expression of gratitude
29
94%
17
56%

Goodbye
--
--
--
--

Money-related utterance
1
3%
1
3%

Service provider repetition of order
--
--
--
--

Gratitude + well-wishing
--
--
--
--

Gratitude + goodbye
--
--
2
6%

Gratitude + well-wishing + goodbye
--
--
--
--

Money-related + gratitude
1
3%
9
29%

Money-related + goodbye
--
--
2
6%

Money-related + gratitude + goodbye
--
--
--
--
Total closing strategies
31
100%
31
100%
No verbal closing
--
--
--
--
Total interactions
31
100%
31
100%
Total = 62 interactions                                                            
Tab. 6: Strategy use in closings (role-play data)


Type
Week


Time 1
Time 2
Time 3

No.
%
No.
%
No.
%
Closing strategies







Expression of gratitude
13
57%
6
40%
11
52%

Goodbye
--
--
1
7%
--
--

Money-related utterance
1
4%
2
13%
2
10%

Service provider repetition of order
1
4%
--
--
1
5%

Gratitude + well-wishing
1
4%
--
--
--
--

Gratitude + goodbye
2
9%
2
13%
--
--

Gratitude + well-wishing + goodbye
--
--
1
7%
--
--

Money-related + gratitude
2
9%
--
--
--
--

Money-related + goodbye
--
--
--
--
--
--

Money-related + gratitude + goodbye
--
--
--
--
1
5%
Total closing strategies
20
87%
12
80%
15
72%
No verbal closing
3
13%
3
20%
6
29%
Total interactions
23
100%
15
100%
21
100%
Total = 59 interactions
Tab. 7: Strategy use in closings (naturalistic data)


In both studies, money-related utterances (e.g., cuatro cincuenta ‘four fifty’) were found in combination with expressions of gratitude and goodbyes. Notably, while the frequency of money-related closing strategies remained stable in the naturalistic data from Time 1 (13%) to Times 2 (13%) and 3 (14%), there was a marked increase in money-related closing strategies in the role-play data over time (Time 1: 6%; Time 2: 38%).

Another difference observed in the two data sets concerned the number of unique closing strategies and combinations of strategies. The naturalistic data contained five different single strategies (i.e., gratitude, goodbye, money-related, server repetition, and well-wishing) and five combinations, whereas the role-play data contained three individual strategies (i.e., gratitude, goodbye, money-related) and three combinations. Two strategy types were observed only in the naturalistic data: a service provider repetition or “play back” of the order and well-wishing (e.g., ¡Tenga un buen día!‘Have a good day!’). An example of the closing phase appears in (3) above: the student’s contribution to the closing includes the money-related utterance Aquí tiene (‘Here you are’) in line 10, and a goodbye (hasta luego) in line 12.

A final difference concerned the absence of a verbal closing phase. In the naturalistic data, 13 % to 29% of the interactions at each time did not have a closing phase. The role plays, on the other hand, always contained a closing phase. In (5), for example, the verbal exchange ends with a request; neither the student nor the service provider used verbal strategies to wrap up the interaction. The encounter presumably ended with the bartender handing the student the two beers that she had ordered[1]. As was the case for openings, the greater variability in the local conditions of the naturalistic interactions and the existence of a real-life communicative context for the interaction, compared to the role plays, may account for the fact that some naturalistic encounters did not have a closing phase:

(5) Samantha (S) ordering at the bar at Time 3 (naturalistic study)
1         S:    hola heh heh  ((laughter))
2         S:    dos cañas, por favor

1         S:    hi heh heh  ((laughter))
2         S:    two beers, please


4.5. Length

A comparison of the length of food- and drink-related service encounters in the naturalistic study (N = 17) and a random sample of the role plays (N=17) indicated that the mean number of turns and words per interaction was higher in the naturalistic data. As shown in Table 8, the naturalistic interactions were approximately twice as long as the role-play interactions, with the encounter being considered as a whole as well as students’ contributions alone. The differences reached statistical significance for the total number of turns (p < .047), the total number of student turns (p < .012), and the total number of student words (p < .027):




Minimum
Maximum
Mean
Std. Deviation
Significance
Total turns






Roleplay data
4
18
8.24
3.750
.047

Naturalistic data
1
48
15.65
13.861
Total words






Roleplay data
4
45
26.59
10.471
n.s.

Naturalistic data
4
126
41.59
36.159
Total student turns






Roleplay data
2
8
4.18
1.629
.012

Naturalistic data
1
23
8.24
5.772
Total student words






Roleplay data
2
28
14.94
6.077
.027

Naturalistic data
4
55
25.35
16.948
Total interactions (role play) = 17
Total interactions (naturalistic) = 17
Table 8: Length of interaction

With respect to range and standard deviation, Table 8 also indicates that the naturalistic data were more variable than the role-play data. The greater range and higher standard deviations in the naturalistic data show that those interactions were more variable in length, with some being very short and others relatively long. In contrast, the role-play data display lower range and standard deviations, indicating that the number of words and turns produced in each interaction was more consistent and less variable.


5   Limitations of the Study

The present study contains some limitations which will be pointed out before the implications of the results are discussed. First, the two research projects on which the present analysis draws were not originally designed and carried out for the purpose of comparing data collection methods. For that reason, some elements of the design of the naturalistic study make for a less-than-ideal comparison. One limitation is the fact that not all of the naturalistic service encounters included in the analysis involved ordering a drink—as was the case in the role plays—a fact that potentially introduces intervening variables. Secondly, although the size of the sample in the naturalistic study was appropriate for the original ethnographic design of the study, for the present analysis the inclusion of only seven participants is a limitation.

Despite these limitations, we believe that the results still offer important insight into the issue of data collection methods in pragmaticsThe results suggest that role plays are a valid alternative to naturalistic data in pragmatics, but confirm previous research suggesting that there exist certain differences between naturally-occurring and elicited interactions which researchers should keep in mind when employing role plays. Moreover, the results provide greater insight into the specific ways in which naturalistic and elicited social interaction may differ in the Spanish service-encounter setting - a topic to be discussed further in the following section.


6   Discussion

The question raised in the present study concerns the extent to which role-play data capture the features of authentic discourse in the service-encounter context. The results point to many similarities between the role-play and naturalistic data. In both data sets, participants overwhelmingly oriented towards the same activities in a relatively fixed order: openings, requests, negotiation, and closing. Turn-taking was realized in a similar fashion in both types of data. Hence, the role-play data were similar to the naturalistic data in terms of structure, sequence, and turn-taking. The negotiation phase was also optional in both data sets and when present, it served the function of exchanging information about price and the needs of the customer regarding the product.

Elicited and naturalistic data also shared similarities with regard to strategy use and development over time. Students in both studies overwhelmingly opened the interactions using greetings, closed the encounters with expressions of gratitude, and employed many of the same head act strategies in their requests. The two data sets also had in common three developmental trends concerning request head acts:

-       a substantial reduction in the inappropriate query permission strategy (i.e. puedo tener),
-       an increase in direct strategies, and
-       a shift away from speaker-oriented strategies.

These similarities in structure, sequence, and strategy use suggest that the role-play data do, indeed, capture many features of authentic discourse in service encounters. However, a closer look also revealed important differences. First, the naturalistic interactions were more variable than the role plays in terms of length and the presence or absence of the opening and closing phases. This greater variability appeared to reflect the reality of local conditions in real-life retail and food establishments, where time pressure, physical constrains (e.g., noise), and the desire to meet the customer’s needs influence the shape of service-encounter interactions. Parties to the naturalistic interactions may not always have found it necessary to verbalize the initiation and termination of the encounter; interactions could be initiated by the presence of the customer at the counter and terminated by the exchange of products and payment. The role plays, on the other hand, were relatively consistent: they were more uniform in length and almost always included an opening and a closing.

Despite greater variability in length, naturalistic service encounters were, on the average, twice as long as the role-play interactions. Both students and service providers in the naturalistic data produced more speech, primarily in the request and negotiation phases, in which more information about products was exchanged compared to the role plays. This difference was a reflection of the effort made by participants to ensure that the customer’s needs were met in the real-world exchange of goods and services.

In addition to the issues of variability and length, there were specific features observed in the naturalistic data which were not captured in the role-play interactions. Two types of closing strategies (i.e., server repetition, well-wishing) and two closing strategy combinations found in the naturalistic data were not produced in the role plays. Non-transactional talk was also present in the naturalistic data, but never in the role plays.

Conversely, the role-play data contained some characteristics that were not observed in the naturalistic data, namely, the overwhelming use of the want-statement head-act strategy, occasional lack of lexical specificity in requests for beverages (e.g., algo de beber (‘something to drink’)), presence of alerters in openings (e.g., perdóname (‘excuse me’)), and relatively high levels of lexical and syntactic mitigation. In the case of the first two characteristics, the description of the role-play situation which the students were given to read immediately before starting their interactions almost certainly influenced their lexical choices with regard to the verb querer (‘to want’) and the phrases algo de beber(‘something to dring’) and una bebida (‘a beverage’).

It is not completely clear, however, why the role plays differed from the naturalistic interactions with regard to alerters and request mitigation. Moreover, this finding is not limited to L2 learners, but also has parallels in the NS data. Spanish NSs, who completed the same role play as the learners in Bataller’s (2008) study, also employed alerters and internal mitigation, whereas in naturalistic service encounters realized by NSs of Peninsular Spanish, Placencia (2005) reports little mitigation in requests and makes no mention of alerters in openings.

In the case of alerters, some role-play participants may have felt the need to use this strategy in order to get the attention of the respective interlocutor before making the request. Indeed, in real-life service encounters, alerters are used for precisely this function: when a customer needs to get an employee’s attention (e.g. because the employee has not seen the customer enter the shop), an alerter serves to get his or her attention and makes relevant a subsequent request for a product or service. The use of alerters by the role-play participants may be explained by the need in a simulated interaction to use language to recreate an imagined scene in the absence of extra-linguistic cues which would establish the communicative context a priori.

The reason why role-play participants more frequently mitigated their requests compared to the students in the naturalistic study may be explained by reference to an argument made previously in the literature (e.g. Aston 1995, Barron 2003, Golato 2003), that is, that the role plays reflected how participants thought  they would make requests or believed they should make requests — based on stereotypes regarding how customers should act — rather than how they actually realize this speech act in authentic contexts.

The results of the present study indicate that role-play interactions captured many of the features of authentic discourse and L2 developmental trends observed in the naturalistic service encounters, while also presenting important differences which can be explained — at least in part — by method effect and the nature of simulated interaction. As has been found in previous studies, role plays have the methodological advantage of comparability, since role-play interactions are more uniform than naturalistic data with respect to length, consistently include openings and closings, and do not involve non-transactional talk. However, role plays have the methodological disadvantage of being susceptible to influence from the description of the scenario and the fact that simulated interaction may produce some seemingly inauthentic uses of language.


7   Conclusions

Naturally-occurring data represents the most direct measure of language use in real-life interactions and should, for this reason, be the preferred method when feasible. However, when collecting natural data is impractical, this study indicates that open role plays are a reasonably valid alternative in the service-encounter context. The findings also suggest that researchers who use role plays would be well-advised to think carefully about the description of the role-play scenarios in order to minimize the possibility that the task description will influence participants’ language use. Another implication of these results is that research based on role-play data would be enhanced by collecting at least some naturally-occurring data, as a means to help to confirm developmental trends and identify those features of role-play data in which the design of the task or simulated interaction itself may have influenced participants’ performance.



References

Aston, Guy (1995). Say ‘thank you’. Some pragmatic constraints in conversational closings, Applied Linguistics,16(1), pp. 57-85.

Barron, Anne (2003). Acquisition in interlanguage pragmatics: Learning how to do things with words in a study abroad context. Amsterdam: John Benjamins.

Bataller, Rebeca (2008). Pragmatic development in the study abroad setting: requesting a service in Spanish.Unpublished doctoral dissertation. The University of Iowa, Iowa City.

Bataller, Rebeca (2010). Making a request for a service in Spanish: Pragmatic development in the study abroad setting, Foreign Language Annals, 43(1), pp. 160–175.

Beebe, Leslie M. & Cummings, Martha C.(1996). Natural speech act data versus written questionnaire data: How data collection method affects speech act performance. In Susan M.

Gass & Joyce Neu (eds.). Speech acts across cultures: Challenges to communication in a second language, pp. 65-86. New York: Mouton de Gruyter.

Billmyer, Kristine & Varghese, Manka (2000). Investigating instrument-based pragmatic variability: Effects of enhancing discourse completion tests. Applied Linguistics, 21(4), pp.  517-552.

Blum-Kulka, Shoshana, House, Juliane & Kasper, Gabriele (eds.) (1989). Cross-cultural pragmatics: Requests and apologies. Norwood, NJ: Ablex.

Callahan, Laura (2009). Spanish and English in US service encounters. Basingstoke: Palgrave Macmillan.

Cohen, Andrew D. (1996). Speech acts. In Sandra Lee Mckay & Nancy H. Hornberger (eds.). Sociolinguistics and language teaching, pp. 383-420. Cambridge: Cambridge University Press.

Cohen, Andrew D. (2004). Assessing speech acts in a second language. In Diana Boxer & Andrew D. Cohen (eds.).Studying speaking to inform second language learning, pp. 302-327. Clevedon, UK: Multilingual Matters.

Crookall, David & Saunders, Danny (1989). Communication and simulationClevedon, UK: Multilingual Matters.

Danblon, Emmanuelle, de Clerck, Bernard & van Noppen, Jean-Pierre (2005). Politeness in Belgium: Face, distance and sincerity in service-exchange rituals. In Leo Hickey & Miranda Stewart (eds.). Politeness in Europe, pp. 45-57. Clevedon, UK: Multilingual Matters.

Drew, Paul & Heritage, John (1992). Analyzing talk at work: An introduction. In Paul Drew & John Heritage (eds.).Talk at work, pp. 3-65. Cambridge, UK: Cambridge University Press, Cambridge.

Eisenstein, Miriam & Bodman, Jean W. (1993). Expressing gratitude in American English. In Gabriele Kasper & Shoshana Blum-Kulka (eds.). Interlanguage pragmatics, pp. 64-81. New York: Oxford University Press.

Félix-Brasdefer, & César, J. (2003). Validity in data collection methods in pragmatics research. In Paula Kempchinsky & Carlos E. Piñeros (eds.). Theory, practice, and acquisition. Papers from the 6th Hispanic Linguistics Symposium and the 5th Conference on the Acquisition of Spanish and Portuguese, pp. 239-257. Somerville, MA: Cascadilla Press.

Félix-Brasdefer & César, J. (2007). Natural speech vs. elicited data: A comparison of natural and role play requests in Mexican Spanish. Spanish in Context, 4(2), pp. 159-185.

Franch, Patricia B. & Lorenzo-Dus, Nuria (2008). Natural versus elicited data in cross-cultural speech act realisation: The case of requests in Peninsular Spanish and British English. Spanish in Context, 5(2), pp. 246-277.

García, Carmen (1993). Making a request for a service and responding: A case of Peruvian style. Journal of Pragmatics, 19, pp. 127-152.

Gavioli, Laura (1995). Turn-initial versus turn-final laughter: Two techniques for initiating remedy in English/Italian bookshop service encounters. Discourse Processes, 19, pp. 369-384.

Golato, Andrea (2003). Studying compliment responses: A comparison of DCTs and recordings of naturally occurring talk. Applied Linguistics, 24, pp. 90-121.

Hartford, Beverly S. & Bardovi-Harlig, Kathleen (1992). Experimental and observational data in the study of interlanguage pragmatics. In Lawrence F. Bouton & Yamuna Kachru (eds.). Pragmatics and language learning, vol. 3, pp. 33-52. Urbana, Illinois: University of Illinois at Urbana-Champaign.

Hassall, Tim (1997). Requests by Australian learners of Indonesian. Unpublished doctoral dissertation, Australia National University, Canberra.

Hickey, Leo (2005). Politeness in Spain: Thanks but no ‘thanks’. In Leo Hickey & Miranda Stewart (eds.). Politeness in Europe, pp. 317-330. Clevedon, UK: Multilingual Matters.

Ide, Risako (1998). ‘Small talk’ in service encounters: The creation of self and communal space through talk in America. Unpublished doctoral dissertation, University of Texas, Austin.

Kasanga, Luanga A. (1999). Pragmatic competence: A longitudinal observation. Indian Journal of Applied Linguistics, 25, pp. 87-104.

Kasper, Gabriele (2000). Data collection in pragmatics research. In Helen Spencer-Oatey (ed.). Culturally speaking: Managing rapport through talk across cultures, pp. 316-369. London, UK: Continuum.

Kasper, Gabriele & Dahl, Marete (1991). Research methods in interlanguage pragmatics. Studies in Second Language Acquisition, 13, pp. 215-247.

Kerbrat-Orecchioni, Catherine (2005). Politeness in France: How to buy bread politely. In Leo Hickey & Miranda Stewart (eds.). Politeness in Europe, pp. 29-44. Clevedon, UK: Multilingual Matters.

Kidwell, Mardi (2000). Common ground in cross-cultural communication: Sequential and institutional contexts in front desk service encounters. Issues in Applied Linguistics, 11(1), pp. 17-37.

Lamoureux, Edward L. (1988). Rhetoric and conversation in service encounters. Research on Language and Social Interaction, 22, pp. 93-114.

Margalef-Boada, Trinidad (1993). Research methods in interlanguage pragmatics: An inquiry into data collection procedures. Unpublished doctoral dissertation, Indiana University.

Márquez Reiter, Rosina (2000). Linguistic politeness in Britain and Uruguay: A contrastive study of requests and apologies. Amsterdam: John Benjamins.

Márquez Reiter, Rosina (2005). Complaint calls to a caregiver service company: The case of desahogo. Intercultural Pragmatics, 2(4), pp. 481-514.

Márquez Reiter, Rosina (2006). Interactional closeness in service calls to a Montevidean career service company.Research on Language and Social Interaction, 39, pp. 7-39.

Márquez Reiter, Rosina & Placencia,  María E.(2004). Displaying closeness and respectful distance in Montevidean and Quiteño service encounters. In Rosina Márquez Reiter & María E. Placencia (eds.). Current trends in the pragmatics of Spanish, pp. 121-156. Amsterdam: John Benjamins.

Omar, Alwiya S. (1991). How learners greet in Kiswahili: A cross-sectional survey. In Lawrence F. Bouton & Yamuna Kachru (eds.). Pragmatics and language learning, vol. 2, pp. 20-32. Urbana, IL: University of Illinois at Urbana-Champaign.

Pinto, Derrin R. (2002). Perdóname ¿llevas mucho esperando? Conventionalized language in L1 and L2 Spanish. Unpublished doctoral dissertation, University of California, Davis.

Placencia, María E. (1998). Pragmatic variation: Ecuadorian Spanish vs. Peninsular Spanish. Spanish Applied Linguistics, 2, pp. 71-106.

Placencia, María E. (2004). Rapport-building activities in corner shop interactions. Journal of Sociolinguistics, 8(2),pp. 215-245.

Placencia, María E. (2005). Pragmatic variation in corner store interactions in Quito and Madrid. Hispania, 88, pp. 583-598.


Rintell, Ellen M. & Mitchell, Candace J. (1989). Studying requests and apologies: An inquiry into method. In Shoshana Blum-Kulka, Juliana House & Gabriele Kasper (eds.). Cross-cultural pragmatics: Requests and apologies, pp. 248-272. Norwood, NJ: Ablex.

Rose, Kenneth (1994). On the validity of discourse completion tests in non-western contexts. Applied Linguistics, 15, pp. 1-14.

Shively, R. L. (2011). L2 pragmatic development in study abroad: A longitudinal study of Spanish service encounters.Journal of Pragmatics, 43(6), 1818-1835.

Sifianou, Maria & Antonopoulou, Eleni (2005). Politeness in Greece: The politeness of involvement. In Leo Hickey & Miranda Stewart (eds.). Politeness in Europe, pp. 263-276. Clevedon, UK: Multilingual Matters.

Tran, Giao (2006). The nature and conditions of pragmatic and discourse transfer in cross-cultural interaction investigated through naturalized role-play. Munchen: Lincom Europa.

Traverso, Véronique (2001). Syrian service encounters: A case of shifting strategies within verbal exchange.Pragmatics, 11(4), pp. 421-444.

Turnbull, William (2001). An appraisal of pragmatic elicitation techniques for the social psychological study of talk: The case of request refusals. Pragmatics, 11(1), pp. 31-61.

Vélez, Jorge A. (1987). Contrasts in language use: A conversational and ethnographic analysis of service encounters in Austin and San Juan. Unpublished doctoral dissertation, University of Texas, Austin.

Ventola, Eija (1987). The structure of social interaction: A systemic approach to the semiotics of service encounters. London, UK: Frances Pinter.

Yuan, Yi (2001). An inquiry into empirical pragmatics data-gathering methods: Written DCTs, oral DCTs, field notes, and natural conversations. Journal of Pragmatics, 33, pp. 271-292.

Zimmerman, Don H. (1992). The interactional organization of calls for emergency assistance. In Paul Drew & John Heritage (eds.). Talk at work, pp. 418-469. Cambridge, UK: Cambridge University Press.




Authors:

Rebeca Bataller, Ph.D
Assistant Professor
Spanish Department
Gettysburg College
Campus Box 411
300 N. Washington St.
Gettysburg, PA 17325
Tel.: 717-337-6860


Rachel L. Shively, Ph.D.
Assistant Professor
Department of Languages, Literatures, and Cultures
Illinois State University
Campus Box 4300
Normal, IL  61790
Tel:  309.438.7185
Fax: 309.438.8038



1.     [1] It is unknown whether the student paid for the drinks immediately or whether, as is common in bars, she paid later on, before leaving the establishment. Some aspect of the specific situation at that moment may have precluded a closing, for example, the bartender may have been talking with another customer when she served the beer to the student.