Dressing and Being : Appraising Costume and Identity in English Second-Language Drama

In many cultures, L2 students are reticent to engage in spontaneous oral L2 production. In Chinese culture, social norms tend to place value on accuracy, which tends to inhibit learners from authentic oral use of the target language. The purpose of this study was to consider the impact of costume, as used in L2 drama, on L2 selves, and attitudes towards specific elements of authentic language use. Costume has long been understood as eliciting imagination, and permitting the expression of possible and desired selves. Fashion ensembles of many kinds are experienced as having a semiotic “sparkle”, which wearers connect to their own self, as they imagine and perform possible selves. In this study, 78 second-language actors were asked to write a brief commentary on how they responded to their costume. This qualitative data was analysed using Appraisal analysis, indicating a majority of positive evaluations. It was also analysed using possible self theory. Comments also showed that L2 actors felt that costumes impacted their emotions and imagination of self, which improved their second language use, cultural performance. They felt costume integrated their oral production with their choices of social register, and their paralinguistic and kinetic performance.


Introduction
Drama is widely understood as an enjoyable language-learning activity.Yet there are few studies of drama as a language-learning vehicle, because few tertiary faculty have expertise in both areas.For this reason, "drama" is treated as a unitary entity, and the contribution of its multiple semiotic and communicative elements lumped together, rather than being independently investigated.This study is the first to consider the contribution of costume, in the context of second-language (L2) learning.Many studies emphasise how positive emotions help learners in the long-term work of acquiring a second language (Cook, 2000).Intrinsic motivation is crucial for investment, and pleasurable activities like drama are intrinsically motivating (Cziser & Dörnyei, 2005).Drama is rooted in the spontaneous fantasy play of children (Galda & Pellegrini, 2008).It scaffolds players in symbolically rehearsing new identities and social interactions (Cattanach, 2005).These aspects of drama immerse L2 players in the target language (Winston, 2011).Drama permeates the L2 curriculum (Taylor, 2000;Heldenbrand, 2003;Elgar, 2002).It facilitates the embodied, personal exploration of texts and contexts (Smith & Herring, 2001), supports oral production, and links oral with written language (Anderson, Hughes, & Manuel, 2008;Miccoli, 2003), and develops realistic cultural performance (Somers, 2000).L2 drama is now found in gaming and virtual environments (MUVEs) (Hubbard, 2002).
Drama engages learners socialised to be passive learners, who are reticent to speak in the target language, such as Chinese L2 students.In Hong Kong, past L2 teaching emphasised "formal features of the language at the expense of encouraging students to use the language" (Education Commission, 2000, p. 25).Hong Kong has recently adopted a Language Arts curriculum.New learning objectives include using English in authentic ways, to gain knowledge of self and others, for pleasure, personal expression and development, in real and imaginary situations (Curriculum Development Council & Hong Kong Examinations Assessment Authority, 2007).Drama is recommended for use (CDC 2002, Education Manpower Bureau, 2005).Many Hong Kong language teachers remain uncertain about how to bring drama into their classrooms.For Confucian learners, language learning is considered useful for future employment purposes (Cheng, 2002).Chinese L2 learners tend to focus on accuracy, devaluing oral participation (Shi, 2006).This study explores costume as an element of drama which helps Chinese L2 learners set aside culturally normative behavior, and immerse in authentic oral interactions.

Literature Review
Costume carries symbolic meanings in many human cultures (Turner, 1987).Anthropologists have noted that, when people dress the part, they act the part (Bell, 1976).In some sense, dressing is being (Damhorst, 1999).Dress communicates identity through socially-conventional signifying structures which enjoy a combinatorial grammar similar to language (Entwhistle, 2000).Dress encodes and displays various social elements including class, gender and material subculture (Barnard, 2002).In the 21 st century, dress is fluid and subjective (Breward, 2003).Our fashions change because dress is "a center of the self, one that reaches out to sense the world and reverberates almost constantly with anticipated and infinitely interpreted meanings" (Garot & Katz, 2003, p. 422).Young people are particularly aware of the construction of identity through dress, having detailed knowledge of its visual, material and procedural symbolisms (Purdy, 2004).They are acutely aware of the meanings of school uniforms, work clothes, the latest jeans, trainers, luxury brands, and their various roles in identity performance, through interpersonal transactions (Crossley, 2001).Thus, identity is changeable in consumer societies, with dress expressing self through ongoing processes of redefinition and re-performance (Eicher, 2000).The "social skin" cannot merely be worn, but must be continually noticed, discussed and evaluated in social interactions (Turner, 1993, p. 486).We "do not simply manipulate appearance like a puppet master controlling some artificial extension of [self]" but "live in and through appearance in ways that are obviously profound" (Garot & Katz, 2003, p. 423).Dress is a means of relating to social groups, enacting work and leisure identities, resisting disliked identities, and performing chosen social messages (Crane, 2000).In Asia, fashion-marketing enhances dress as a site of renegotiated identities (Steele & Major, 1999).
In Asian contexts, dress is associated with playful identities (Hancock, 2000).Spectacular dress ensembles, popularised by the media, have accustomed people to the construction of extraordinary selves through a bricolage of elements, including in cosplay (Winge, 2006).Asian L2 learners are familiar with costumed interactions through favorite anime and manga characters (Poitras, 2001).Anthropological studies of cosplay use Huizinga's (1955) spectrum between spontaneous unstructured play (paidia) and rule-bound, goal-oriented games (ludus), and understand costume as masking the real self in order to enable the realisation of a different self (Sutton-Smith, 1997).Cosplay actualises private selves as distinct from public selves identified by work dress or school uniforms (Miller, 1994).Participatory cultures have grown up around cosplay, with costumed fans interacting as characters (Jenkins, 1992).Asian L2 learners understand cosplay as a permissible escape from the discipline of mainstream society, through managed, narrative-based, hyper-real performances of fantasy selves in which individuals perform for themselves as much as for audiences (Lamerichs, 2010).Cosplay in digital contexts involves editing avatar appearance to perform a virtual imagined self (Taylor, 2003).Avatars "allow new configurations of self that fulfill wishes and fantasies" (Veletsianos, 2010, p. 578).The relationship between the avatar and the user is "not stable and unitary, but dynamic and multiple", as the user tries out various selves and performances (Schultze & Leahy, 2009, p. 3).Computer-mediated, game-based language activities are widely used in L2 teaching, throughout Asia (Wallace & Maryott, 2009).Creative control of identity play facilitates language learning, because L2 interactions are required, and feedback develops L2 abilities (Bailenson, Beall, Loomis, Blascovich, & Turk, 2004).
Costume, then, is about self, as negotiated in pleasurable social interactions with others (Welters & Lillethun, 2007).Costumes, as ensembles of elements which combine and layer aesthetic and social meanings, offer desired clusters of meanings intended to excite viewers (Calefato, 2004).Wearing the ensemble and experiencing the responses of others offers players an experience called "heteroglossic sparkle" (Chandler & Chandler-Smith, 2005).Through the interaction of display, viewing and response, costume evokes psychographic responses in the wearer, including body-and gender-perceptions, movements and gaits, attitudes and values, behaviours and social dialects (Benstock & Ferriss, 1994).Barthes described theatre as a semiotic practice notable for its density of signs and information polyphony, coordinating language, gesture, kinesis, and costume (Barthes, 1983).Directors of theatre productions, aware of the meanings associated with specific dress items, select ensembles to encode social identities and relations in order to communicate culturally specific meanings to audiences (Oklobia & Bakare, 2009).Professional actors understand this also.But L2 players are not trained to take this approach.For them, costume remains a pleasurable, playful process focused on self, closer to the fantasy play of children or the creative control of an avatar.
The human tendency to experience dress as a playful ensemble of visual symbols connecting to self can be used in L2 teaching.Self is usually understood as flexible and multiple (Mansfield, 2000).L2 motivation studies have recently been re-theorised using a well-established research paradigm from psychology, self-discrepancy theory (Dörnyei & Ushioda, 2009).This indicates that L2 learners imagine various possible selves, including everyday, ought, desired and possible selves (Schunk & Zimmerman, 2007).The real, or here-and-now self manages practical matters of everyday life.The ought self reflects authoritative voices such as family and teachers, whose messages about what we should be like have been accepted by the here-and-now self (Higgins, 1998).These are messages such as "I should learn to speak English".We also imagine desired selves, composed of personal dreams and wishes.These reflect our chosen view of ourselves, and undergo constant revision and change (Markus & Nurius, 1986), for example, being an excellent dancer, or a wealthy businessperson.When our various selves do not cohere well, we feel an uncomfortable sense of discrepancy.To reduce the discomfort, we seek to reconcile our various selves (Ushioda, 2006).Specifically, we imagine possible selves, composed of experiences and opportunities arising in daily life (Boyatizis & Akrivou, 2006).Possible selves may become "self-guides", helping us make effective decisions to reduce discrepancy, and move towards greater integration of selves, which occurs under two conditions.First, visual elements of the imagination must be richly detailed.Second, tasks set for the possible self must be realistic for the here-and-now self to attempt (Dörnyei & Ushioda, 2009).When learners can control the possible self's performance of the imagined behaviour, they increase their here-and-now competence.Over time, this possible self becomes incorporated within the here-and-now self (Hoyle & Sherrill, 2006).This occurs best with intrinsically motivating activities such as drama (Sandrock, 2002).L2 learners are notable for the ability to imagine possible selves, from constantly imagining possible L2 selves in role-plays and other classroom activities.Through processes of aspiration and revision, they maintain multiple imagined local and international, personal and professional identities (Yashmina, 2002).
This study explores how costume facilitates L2 students' imagination of possible English-speaking selves, and supports their performance of them.It considers how costume motivates Chinese L2 actors' imagination of possible selves, which function as self-guides.Research questions used in this study were: how do Chinese L2 students respond to costume in L2 drama?How do their comments connect costume to the imagination of possible selves?What does this mean, for the use of drama in teaching English to Chinese L2 learners?

Method
Data was taken from 78 L2 actors with speaking roles in one of five English-language play performances produced over a 5-year period.All were native speakers of Cantonese or Putonghua, enrolled in tertiary English programmes, having gained a proficiency of IELTS 6.5.Chinese-speakers with non-speaking roles, non-Chinese speakers with speaking roles, and Chinese-speakers with extensive prior theatre experience were excluded from the data.All actors experienced about 60 hours of rehearsal over a period of 7-10 weeks, followed by 5-8 live performances over 3-4 days.
Costume was integrated into rehearsal from week 4. Actors chose their costumes in consultation with the group, from a collection of hundreds of professional pieces covering many cultures and periods.As the purpose of the performances was to communicate the play's meaning to L2 student audiences, no effort was made to teach authentic period costume, either to actors, or audiences.While actors were pleased to participate in classic dramas, by authors such as Shakespeare, playtexts were substantially cut and completely modernized.Actors often combined historical with contemporary dress, for example wearing Elizabethan breeches and hose with Doc Martens and extreme dance-club makeup.Actors updated traditional costumes, for example slashing open the front skirt of a traditional ball-gown to reveal hot pants, fishnet stockings and knee-boots.Thus, they were exploring costume as visually meaning, rather than reproducing a target culture or period.Some productions were more visually heterogeneous than others, depending on the actors involved.The production of Pride and Prejudice was visually homogeneous but historically inaccurate, as the actresses preferred the close-bodied, open-skirt gowns with full trims, frills and petticoats, a style which pre-dated Austen by 60-80 years, to the historically accurate but simpler Empire dresses.Authenticity in dramatic performances has tended to be the preserve of companies playing to wealthy classes within the cultural industries (Rumbold, 2010).Contemporary English theatre performance was reinvigorated by incorporating pop culture elements, usually attributed to Peter Brooke's 1970 modern-dress Midsummer Night's Dream.It is now quite usual to mix contemporary with historical elements (Dessen, 2002).L2 learners have little need of a historical understanding of culture or costume, where they are motivated by taking on the semiotic power of traditional garments and meanings, and mixing these with frankly local, glamorous and contemporary dress items and meanings.This was the approach taken in all five productions.It was left up to actors to interpret these combinations of costume elements, as they developed their roles.
Data was collected one week after performances finished.Participants were asked to write a brief response to questions covering several aspects of their experience.Two addressed costume.Questions used simple grammar, connecting major semantic elements to personal experience (Block, 1998): "Was costume important to you as an actor?" and "How did your costume affect you as an actor?"These were intended to elicit fulsome comment.The first was framed as a quality ("important"), the second as an experience ("affect") for evaluation.Data was aggregated for analysis.Data was analysed using Appraisal analysis, which details how attitude is encoded in language by taxonomising realisations of subjectivity (Martin, 1995).When expressing personal opinions, people choose from a large number of possible lexicogrammatical options (Martin & White, 2005).Appraisal sorts choices into system networks constructed from the options available in the English language (Macken-Horarik & Martin, 2003).Text corpora can be analysed, using the three system networks for affect, judgment and appreciation (Halliday, 1994).Networks are articulated, as in Figure 1.Semantically, L2 learners talk about imagined possible selves the same way they do the here-and-now self (Dörnyei, 2005).Tabulating choices within the Attitude networks affords a detailed picture of how L2 learners respond to costume and connect them to possible selves.Appraisal is widely used in sentiment analysis, as is the use of software to tag and sort lexis into hierarchical classes is a well-established method (Wiebe, Wilson, & Cardie, 2005).Appraisal networks are now a routine part of computational linguistics (Scherer, Schoor and Johnstone, 2001).This study used the software CorpusTool (CT) (O'Donnell, 2008).While attributions can sometimes be ambiguous, studies of both human and automatic tagging indicate their reliability (Taboada & Grieve, 2004).Attitude is the easiest to attribute (Read & Carroll, 2010).The validity of attitude tagging is strengthened by its consonance with attitudinal systems used in psychology (Oatley, Keltner, & Jenkins, 2006)."Appraisal theories of emotions have gained widespread acceptance in the field of emotion research" (Bednarek, 2009, p. 150).
This study focused on directly realised ("that wig was hilarious") and indirectly invoked ("that butler suit never fit right") realisations (Bednarek, 2006).Qualitative data from student actors was digitised and uploaded to CT.
Layers specific to analysing attitude systems were created.Attitude attributions were reviewed by two human taggers, both with more than 200 hours' experience.Prior to review, both taggers shared a 3-hour norming session, jointly analysing examples and comparing results.Taggers then checked attributions independently, after which the inter-rater reliability was calculated using Cohen's κ (Lombard, Snyder-Duch, & Bracken, 2004).This was 0.915 (p-o) and 0.903 (f-m).Both values are stronger than we would attribute to chance.

Discussion
This data suggests costume is experienced by L2 learners as special, attractive in quality and impact, and tending to create a sense of competence, enabling a sense of veracity, and creating feelings of happiness, confidence and interest.Negative realisations were found in four of the same subcategories as positive, suggesting these are important areas for L2 drama: the impact and quality of the costume, how much the actor likes it, and how special it seems.Participants responded negatively when they found their costume boring or ordinary, when they disliked it, or when it was unfamiliar.While negative evaluations were few, they revealed that L2 actors relate costume to self, as they responded negatively when a costume did not make them feel exceptional, or did not create a desired impact.For example, one actor's comment on her role in Jane Austen's Pride and Prejudice was, "Even I am a Bennett girl, that apron makes me boring, I hate it".This negative appreciation of the apron relates it to the actress herself.She is not considering its historical verisimilitude, or the visual coherence of the Bennett girls as an onstage ensemble, things which directors must consider.
The low frequency of negative realisations and the absence of realisations of fear-misery, insecurity-disquiet and dissatisfaction-ennui suggest the motivating power of costume.Most comments showed actors focused on their own responses to costume, implicitly assuming that their own perception was shared by cast members and audiences.That is, they assumed that the semiotic sparkle they felt from the costume was the sparkle they performed and audiences saw.Overall, this provides initial support for the notion that costume is important to culturally reticent L2 actors, and can intrinsically motivate them in oral production.
We can see this more easily, by exploring some examples, which show how costume impacts emotions, promotes imagination, connects actors to possible selves, assists them in the oral and social performance of their role, affects L2 use, and increases awareness of communicating to an audience.

Realisations about Costume Impacting Emotions
Many participants (48=61.54%)described how costume created a sense of sparkle for their here-and-now self, and viewed their emotional reaction to this as important to their acting.
1) It makes me feel more excited.
2) Very often costume makes my emotions and feelings are shown.This really counts.
3) I become emotionally love drama.Love this! 4) It is really wonderful to dress in this beautiful clothes.I am so happy to be like Jane Austen heroines.
5) I look so excellent, I love this dress.I am so impressed to see my acting in this.

6)
Costume not only has the function to convey the character, the status, the message for the actor who wears, it also creates magic when it has different combination, different matching.For example, I am changing my costume and the new dress matches with different accessories, it creates different images.
These responses realised appreciations of reaction-impact ("excited", "wonderful"), feelings of satisfaction-pleasure ("impressed") and happiness-cheer ("love", "happy"), consonant with contemporary theoretical understandings of how we "live through" dress a way of constructing a desirable self.Participants reported that costume changed the here-and-now self ("I become", "I am so impressed to see my acting in this").
Costumes were described as causing these changes ("makes my emotions and feelings are shown").Example 6, written by an actress who had a costume change midway through the performance, articulates her experience of both dress ensembles as distinct semiotically ("convey…the message", "creates different images"), and experientially transformative (creates magic").

Realisations about Costume Promoting Imagination of Selves
Many participants (43=55.13%)described costume as productive of imaginations which helped the actor interpret their role: 7) The costumes is so beautiful and I'm full of imaginations once I get dressed.
8) When one puts on the costume, he is aware that he is taking up a role.No longer is he the real self, so he immediately acts including how he walks or speaks.An interesting experience! 9) When I dress up as a wealthy man, I feel more confident and walk confidently.Here, actors described costume as producing imaginations ("full of imaginations"), causing changes in here-and-now self-image, and eliciting possible selves ("no longer … the real self", "can create characters", "new image", "feel my part", "It makes a totally different image").This included changes in physical self-image ("beautiful"), and physical performance ("movement", "walks" twice, "movement, gesture, posture.Enhances my character").Participants noted that imaginations created by costume impacted their L2 performance ("how he … speaks"), and reduced discrepancy ("feel more confident", "confidently").The use of "one" and the third personal pronoun "he" to generalise a personal experience (example 8) was found in only 16 sentences (2.17%).
Most responses used personal pronouns.

Realisations Connecting Costume to Possible Selves
Many participants (40=51.28%)selected processes connecting costume to possible selves: 16) It figures us into different roles, and even a few differences is able to make us changed.17) When I wear the costume, I think that I've changed my personality.With different clothes, your idea and even your movement will change.
18) Enhancing by creating a different personality.These clothes can stimulate me to know the character more.
19) After putting on different costume, suddenly I can identify my character more clearly, and personality becomes more outstanding and easier to understand.20) Change some different things can totally change the character's personality, even you have never been that before.
21) It lets me see the social status of the character.So, it changes the way I walk & use other physical expressions.
22) It focuses my character from self consciousness and I adventure into the new role.
In 16, 19 and 20, the modality of possibility ("can", "able to") and related expressions realise a positive ability or potentiality of the self which has been activated by costume (DeCoursey, 2013).In 18 and 19, the actor attributes changes in self to costume.In 20 the counterfactual ("even [if] you have never been that before") and in 19 the perception process ("let", "easier to understand") show the actors understand costume as facilitating imagination of possible selves, via the physical self.Costume is understood as an agent, changing the self ("figures us into", "make us changed", "become outstanding", "change" twice).The role of costume in moving here-and-now self into the role as an imagined possible self is articulated in 20.

Realisations Connecting Costume to Cultural Performance
Some participants' responses (37=47.44%)experienced costume as suggesting social elements of imagined possible selves: 21) Different costumes make different appearances which show different levels and standards.My coat jacket can change my status, when I am the fool and when I am the priest.
22) I wear the lower class dress.I am Maria, she is a bad girl, she's clever and funny.It's one of the most important elements in drama performance, we learn to dress, we learn to act, we learn the role.Some costumes have a strong sex appeal.It is quite demanding for behaviours.
23) Costume presents class, and many different aspects of personality.My costume can help to tell my status, occupation and relations with all the people in Olivia's house.
24) When I dress in these clothes it seems that I am a glorious woman.What I behave is so elegant.I love this colour and I also happy to take pics with my classmates.It is really an unforgettable memory.I will keep it in mind.I like that I am a glossy girl when I talk with other glorious girls.I will keep in my heart forever.
25) What difference the costumes making?I really wanna perform in my costume.I behave differently when I'm in my costume.I can gender-cross in my costume.It help you to get into the character.The majority of us have to act into the opposite gender.Before you change the costume, you're a girl, and say something like a girl, but when you have that costume, it helps you to merge into that character, and even some makeup helps.
26) In the play, the characters are supposed to be a man.But we only have 1 boy so we have to turn the female actors to be men.Without the wigs and the male costumes, we still act a bit girlie.But as soon as we change the costume to be a man, right away I can see the gesture, the walk, the way we talk in a different voice, we cut off the girlie thing.It's more convincing.
27) Costume art adds personality to my character.It helps my character to perform a right trait in the play.Only a little thing, just shoes can change my boldness.In costumes, I think we can really be a king, a queen.
28) The tribal leader's costume looks mysterious and authoritative to me, especially when I put on the mask with angles.I looked handsome and smart and noble after I have changed my costume.I put on my hat and turn much more barbarian.The body gesture and movement changed with my costume.Once I become barbarian I perform much more wildly.
Actors understood costume as expressing elements of identity, including class or status (21,22,23,27), gender and sexuality (22,24,25,26).They realised costume as an agent, with verbs taking actors and dramatic meanings as the grammatical objects of costumes ("make different appearances", "show different levels", "presents class", "tell my status", "help to tell my status", "help you to get into the character", "help you to merge into that character", "helps my character perform a right trait", "change my boldness", and "turn much more barbarian").They connected elements of cultural performance to their here-and-now self, describing acting as behaving, seeming, resembling, wanting and thinking -processes which connect closely with self.They did not choose verbs used for professional actors -perform, stage, present, take the part of, star as, appear as -which highlight rather than close the distance between the person of the actor and the personality of the role.This is seen also in the pronoun confusion ("I am Maria, she is the bad girl"), and in the effusive repetitions of "I" ("I am a glorious woman", "I am a glossy girl").Example 22 describes "drama performance" as beginning with costume ("dress"), extended in acting and learning the role, and evaluated as "important" and "demanding".In 28, when the here-and-now student self dons a "mysterious", "authoritative" costume, its attributes are transferred to him, helping him produce a "handsome", "smart", "noble", "barbarian" (positively evaluated) possible self, which changes his physical performance ("wildly", positively evaluated).In 24, costume enhances social presence ("What I behave is so elegant").The actress has learned to control a desired communicative ability, realistically.This desired self is in process of becoming a viable self-guide, increasing here-and-now competence.

Realisations Showing Costume Impacting L2 Use
Some participants (35=44.87%)noted the impact of costume on L2 usage: 29) Change of costume means I change role, means I change of character, mood and even language.
30) When I wear the costume, it happens that the character changes language, manner and posture to match my new identity.
31) Costume makes different kinds of languages to be borne in mind.With the change of accessories the modification is concretized more dramatically.It's a quick way to vary the register variables.
32) They affect the actors, make them feel more confident in speaking.
33) The costume gave voice to us.All the actors who were dressed up appeared strong as a character.Their dresses made their personalities graceful, elegant and very artistic.It gave confidence to all of us, even some of the quieter and shy people were shining in their dresses.They were naturally using actions, expressions, and they were saying lines and with facial expressions which went very well with their dresses.34) When I was speaking with the actors and actresses it seems to match in accordance with my costume.
35) The costume made the lines between the two of us positioned me to behave as if we are of the same class.
Actors noted that costume changed the way they spoke (29, 30), in the areas of register (31), L2 confidence (32,33,34), expressivity and coordination of language with paralanguage (30, 33), and L2 interactions (34, 35).These impacts were seen as part of a coordinated package including role, emotion and body language, which costume supported and integrated.

Realisations about Costume as Communication
Some participants (29=37.18%)saw costume as part of communication with audiences: 36) Dress can reflect our personality.Different role put on different dress.The audience can be instantly informed about our characters.Effective language learning involves awareness of the recipient of L2 communication.Example 36 shows the naïve actor's view of the identical impact of costume on performer and audience, but includes awareness of the audience.Example 37 moves from seeing costumed actors from the outside ("makes the actors look like"), to the interior experience of the actor ("who we act"), and back to perceiving the play from the audience's point of view.This switching around offers insight into how the L2 actor develops an understanding of reception.We can also see the process of naturalisation of the possible self's competencies with the here-and-now self's normal repertoire.Examples 38 and 39 characterise costume as a communicative medium available for use by the actor, 40 and 41 see it as enhancing L2 communicative proficiency, with reception mirroring the actor's experience.Dress as a communicative medium is evaluated as important, in 42.

Conclusion
This study has produced three main results regarding the use of costume with Chinese L2 drama students.First, the Appraisal data shows that Chinese L2 students evaluate costume positively, reporting that it makes them feel exceptional and capable, happy, confident and interested, and they appreciate the quality and impact of the experience.They respond negatively on some of the same grounds, particularly the impact and quality of the costume, if perceived as boring and ordinary.Second, Chinese L2 students experience costume ensembles as having sparkle, which helps them feel more confident speaking, more able to select L2 registers, and more able to integrate their cultural and paralinguistic performance.This suggests that, to benefit from sparkle, teachers should involve students in the selection of costume items.Further, care should be taken to treat costume playfully, and valorise some kinds of characters through dress, for example villains, worthy but unspectacular characters such as grandparents, and characters of lower social status or social minorities.The motivational value of sparkle may also be used for L2 students taking non-speaking roles, and for L2 crew, and suggests that students should not be pressured to participate in performance, as sparkle motivates intrinsically.
Third, Chinese L2 student actors relate the sparkle of extraordinary dress ensembles to the self.Lacking a professional's perspective of the organised visual semiosis of dramatic performance, they feel that their costume helps them understand and express their role, and that audience receive it similarly.The connection promotes imagination of possible selves, some which can support them in achieving realistic language tasks, including L2 oral production.This connection can be used in classroom activities, as well as performances.Costumed activities could invite students to playfully explore elements of oral performance such as social register.As costume motivates intrinsically, for such activities to be effective they must generate imagined possible selves which L2 learners find desirable, meaning activities must be lightly structured, resembling the participatory cultures of cosplay, where costume reflects fantasy and leisure life rather than the conformity of school discipline.Interactions should remain close to paidia, unstructured play, rather than rule-bound games, or ludus.When producing full performances, teachers can engage L2 students to adjust the language of the script, speak expressively, and integrate gesture and movement consonant with the students' perceptions.
These results have implications for cost, classroom management, and the choice of activities or performances.
For costume ensembles to be perceived as having sparkle, for students to invest in drama, and for their language to benefit from the motivating effects of costume, ensemble items must have emotional impact.They must be exceptional, compared to daily dress.As one participant wrote, "If there wasn't any costumes, can you imagine what would happen?Nothing!It's just like you dance with a naked guy.Nobody wants to see that."Building a costume collection requires a budget.While a collection may be built over some years, the constant renegotiation of identity in consumer societies means that new items must be purchased regularly.Costume collections also have implications for storage, classroom praxis and facilities management.Many L2 teachers cannot access changing-rooms, and cannot allocate time for students to change.This need not limit the use of costume in drama activities, if costume items are selected for being compact, appealing and portable, for example hats, capes, tunics, wigs and hand-held accessories.
This study has limitations.First, the data set was small.This is a consequence of the demanding workload and investment of time required to produce L2 drama performances.These make it difficult to acquire large datasets.Perhaps for this reason, there are no other comparable studies, and the current study has made a substantial contribution to the understanding of the use of drama in L2 studies.At the same time, few tertiary faculty undertake full productions.Given this inherent limitation to our understanding of drama in L2 learning, further research could be undertaken on the basis of single productions, using a case studies approach.Second, Hong Kong is known for its fashion industry, perhaps encouraging a susceptibility to the semiotic sparkle of costume ensembles.Chinese L2 students from other Asian countries may respond differently, realising different attitudes and relating it differently to the imagination of possible selves and L2 oral performance.This is an area where further research is required.Third, the meanings attributed to costume remain indeterminate and changeable.To some degree, the sparkle L2 students experience, and its impact on their imagination and motivation reflects the consilience between specific clothing items chosen, and normative local and contemporary visual interpretations.These are fluid, reflecting popular culture and fashion trends.Thus, it may not be possible to articulate relations between specific costume items or ensembles and L2 attitudes realised, so as to itemise collections or detail budgets.The arbitrary nature of sparkle limits our ability to define what it is about costume that motivates, promotes the imagination of possible selves, and supports oral performance and confidence.At the same time, theories of self and linguistic approaches to subjectivity offer a vehicle for defining personal responses and evaluating the contribution of costume to L2 learning.
Overall, it is clear that L2 students respond imaginatively to the use of costume in L2 drama.In the case of students who tend to be reticent, such as Chinese students, costume offers them a culturally normative channel for self-expression.This allows them to set aside social behavioral norms, such as the concern with accuracy, in favor of creative oral production and L2 language use.
Figure 1.The attitude system 10) Costume changes my movement, gesture, posture.Enhances my character.11) Costume can create characters.It is really exciting and fantastic!12) The first time for me to try to put on the wonderful costume, it help me to create a new image that I never predict before.I like the new image, and I love the drama and totally change my mind to love the character and the drama!13) Put on the dress, I have a feeling that I live another century.My costume helps me feel my part.It makes a totally different image of a character.

37)
Costume makes the actors look like who we act.It can give the audience to follow and make them easier to understand the background of the roles.It gives colours to the play.38) Costume can help me tell the story and give clues for audience to understand the play.39) Costume is essential key to help highlight the theme for the scene.It changes how the audience perceives my character.Just a piece of accessory, even the little piece of black fabric will do the trick.40) If I have costume, I think I act much better.The audience understand what I am doing, my character.It helps audience to understand what the character is like.41) Costume is important if you want to fully immerse or immerse the audience into the story.42) It is pretty important because it gives the audience an impression on what you are doing.The dress gives an impression or the imagination that they are in that world.