Chapter 1 An Introduction to Drama in ELT
The use of drama in ELT (DELT) is not unprecedented, though the applied linguistics literature is scant. An exhaustive search through prevailing texts by noted authors describing language pedagogy such as Brown, Nunan, Harper, Lightbown and Spada, Gass and others reveals next to nothing concerning its use or potential in language acquisition even though modern language textbooks often present short dialogues in text and audio which teachers may use as teaching material. “Under-exploited”, to use Almond’s vernacular (Járfás, 2008), in ELT, there is little tangible evidence to support the use of drama as an effective language teaching strategy, methodology or approach beyond introductory situational dialogue presenting new vocabulary and grammar.
Researchers, teachers and teacher-researchers using drama as a teaching and learning medium report success in its use to engage students effectively in ELT, yet the soft language of the observations may appear unpersuasive to those teachers desiring measurable outcomes from such approaches. To the question of how drama could contribute to cooperative language learning Járfás responds, “drama bears in its nature some cooperative elements” and lists group performance, relaxed atmosphere and mutual respect as the key dignitaries of cooperative language learning through drama. Hall’s (1982) drama-based course attempted, ”within the student’s own parameter of competency in [English], to practice those skills which [the student] had acquired in other courses with a view to reinforcing situational structures already learnt in non-dramatic contexts”(146), with a “definite increase in motivation…observed as the semester progressed”(156).
Language and culture appear to be inextricably linked, yet the subtleties of non-linguistic cues to express emotion are typically ignored in regular conversation courses (Miccoli, 2003: 123). Drama as a language-teaching medium allows learners context to explore these concepts. Miccoli’s own course in a Brazilian university, “English through Drama” underscores these points, establishing character roles and “creating a genuine purpose for communication…a reason to use language” in “meaningful contexts”. Drama provides a natural context for integrating the “four skills”—reading, writing, listening and speaking (Elgar, 2002), allowing for collaboration, negotiation of meaning and the creation of authentic texts, the results of which are “encouraging”.
In a pro gradu DELT thesis, Salopelto, focusing on intercultural competence, notes teaching English intercultural communication in the Finnish local curriculum through contrastive analysis will have questionable results (2008, 20), preferring instead a drama-based approach to develop empathic response in the target language, English. Salopelto further notes that such an approach is time-consuming and may lead to lower proficiency in target language forms (73). This point is picked up by Stinson (in Anderson et al, 2008: 210) whose own study initially focused on “structured activities in which the students needed to use spoken English”, “meaningful talk in the specific dramatic context”, to “improve language outcomes” (ibid, 195; emphasis, author). The study’s results lacked more concrete measurements on the students’ language outcomes save “significant improvement in oral communication”, noting the “unanimously enthusiastic” response of teachers and students, increased confidence in spoken English communication, greater enjoyment of lessons and improved racial relationships in classes.
The studies, above, were performed to confront perceived shortages in ELT—namely motivation, intercultural communication, negotiation of meaning and affect. As such, they are lacking in specific detail towards the nature of language learners’ linguistic improvement. A teacher assessing whether or not to use a dramatic approach to language learning would appreciate the improved mood of the students, increased motivation, negotiation of meaning and collaboration, but the students would still be measured by test results; the literature on DELT glosses over this crucially persuasive information. This is not to override the results of the above-mentioned studies, but, unfortunately, sentiment doesn’t seem to satisfy educational regulators though application of the above studies’ methods and approaches may produce more cosmopolitan citizens.
In the current study, the aforementioned communicative elements of language will be important; linguistic and paralinguistic measurements will be taken through pre- and post-research interviews and texts from a group of thirty university English students split between two English research classes—a drama-based course (D-Class) and a traditionally organized course (T-Class). Through the study, it is hoped that the following questions (and their inferred hypotheses) will be answered:
- What effect will a drama-based approach to language learning have on the development of student narratives?
- Will a drama-based approach to language teaching and learning have a positive effect on student-produced narratives?
- Will EFL students in a Korean university context respond better to a drama-based EFL class environment or a more “traditional” teacher-fronted classroom?
- Will a drama-based approach to language teaching and learning correlate positively with more than reduced negative affect and anxiety?
- Will students who participate in a drama-based language class show more improvement in the production of identity narrative post-learning than those who participate in a “standard” teacher-fronted class?
- Will a drama-based language class allow students to develop more physical, gestural and vocal expressiveness than a class which is more “traditional” in nature?
- What will be the nature of the differences between students participating in a drama-based course and those participating in a traditional teacher-fronted class post-experiment?
- To what extent will the effects of drama-based instruction of metaphor in narrative compare with the effects of explicit instruction of metaphor in narrative construction in a traditional teacher-fronted EFL class?
The raw materials will be compiled into a multimodal corpus where the effects of the research classes can be more easily examined. It is hoped that facts of language, rather than ideas about it (Sinclair, 39) will become evident, both linguistically and paralinguistically. Furthermore, it is hoped that the D-based class will show similar if not greater improvement in both areas, the nature of which will be shortly discussed in chapter two.
Chapter 2 A Review of Linguistics and SLA Literature Underpinning DELT
In this chapter, we will survey the literature contained within various fields of linguistic and SLA inquiry concerning the development, implementation and general use of a DELT approach to language learning. We shall begin with a brief mention of research methods, then move into the psychological/ linguistic branches most important in a DELT approach. Following this, we shall survey the literature surrounding multimodal corpus linguistics, metaphor, and narrative. A discussion about DELT proper will be held in the chapter following. The term approach is here used since it is generally recognized that methodologies are frequently developed with the view that language learners are deficient (Richards and Rogers, 2008: 15). A method is the implementation of decisions made about skills, content, order of presentation and technique to be used (Richards and Rogers, 2001: 19). An approach, conversely, is better considered as flexible and interpretable to a particular learning and teaching situation and may not lead to a particular “method” (ibid, 24; see also Brown, 2007b: 14). What this means for a particular DELT class is the needs of the student at any particular time should constantly remain in view and materials should be ready to move in any number of directions. More will be said on this later, particularly in chapters three and four.
2.1 Research Methods in Applied Linguistics
In the introduction it was noted that the type of research was important to determine the applicability of the results to the ELT field. Three types of research study important for applied linguistics are those which are qualitative, those which are quantitative and those which are mixed-methods, blending elements of both qualitative and quantitative. Each type of research comes with its own ideological difference. Quantitative research tends to produce numerical data from which results are statistically generalizable. In this form of research, it is suggested that research is researcher-controlled, objective, done for the purpose of the data, and is somehow divorced from the researcher to produce generalizable objective “facts” (Nunan, 2008: 3). Qualitative research produces non-numerical data analyzed in a non-statistical way (Dörnyei, 2007: 24). This form of research is relativistic, assumes a subjective stance to the data and allows for non-generalizable results (Nunan, 2009: 3). A mixed-methods study uses both kinds of data. Much work must be done to identify which categories and figures are to be collected before the research is started in a quantitative study; for a qualitative study, much of the study is “emergent”—open to change during the study’s progress (Dörnyei, 2007: 33-37). Mixed methods allow for qualitative and quantitative data to symbiotically support one another (Miles and Huberman, 1994: ibid, 42).
The majority of DELT papers surveyed above tend towards the qualitative. The main weaknesses, while being more flexible, making sense of complexity and having an exploratory nature, are the sample sizes are often small, researchers may directly or inadvertently bias the data, and a preference for substance (research topics) over form (methodology) (Seal et al, 2004: ibid, 41). This study uses a mixed methods approach to formulate conclusions via corpus analysis of student-produced text (verbal—i.e.: linguistic, pitch and intonation; visual—i.e.: gestural, expressiveness; and textual—i.e.: linguistic). It will take advantage of the strengths of both qualitative and quantitative analysis, attempting to describe the complexity of the environment in which the participants are working as well as the nature of the linguistic differences between pre- and post-research interviews, suggesting reasons for noted differences.
The next section will identify potential mechanisms working for or against the observed responses to the research which have come about, giving some explanatory perspective.
2.2 Behaviorism as a Locus of SLA
Drama as a medium for language instruction has three primary linguistic coordinates localized around behavioural psychology, cognitive linguistics and social constructivism. In a behaviorist/ structural linguistic view of SLA, the external environment of the learner is the primary focus, providing the stimuli for responses which, if reinforced, form correct habits—in other words, “learning”, in a behaviorist teacher’s vernacular (Johnson, 2004: 10; Brown, 2007a: 27; see also Lightbown and Spada, 2006: 10). Behaviorism, attempts to answer the question of how languages are learned with the view that imitation of language produced around a learner (in language acquisition theory, the first language learner), receives positive reinforcement if the learner responds to stimuli correctly and in the correct context. Positively reinforced responses which are practiced further and which continue to receive positive reinforcement become habits.
The behaviorist theory of language, posited by B.F. Skinner, was critically received since it seems difficult to conceive that any novel statement an individual could make would be the result of positive reinforcement (Brown, 2007a: 27). Behaviorist research strategies began to fall into disrepute as it was felt subjects were objectified and their own thoughts and feelings about the experiences of the experiment behavior were considered to be irrelevant to scientific inquiry (Johnson, 2004: 10). This was, some thought, potentially dangerous—a researcher’s intention can manipulate the subject’s behavior and individuality of the subjects is ignored (ibid: 11). The Audiolingual Method (ALM) and the Total Physical Response method are two examples of teaching in the behaviorist tradition. Both rely on stimulus and response conditioning for language development (see Brown, 2007a: 111; see also Lightbown and Spada, 2006: 146).
2.3 Cognitive Linguistics as a Locus of SLA
In cognitive psychology, meaning, understanding and knowing are of the utmost importance, and psychologists rely on tools of logic, reason, extrapolation and inference to explain behavior (Brown, 2007a, 11). Rather than trying to describe behavior as behaviorists do, cognitive psychology attempts to explain it. Among the proponents of a cognitive view of linguistics is oft-cited generative-transformational linguist, Noam Chomsky, whose approach, to propose a language acquisition device (LAD) includes a universal grammar (UG) which every child uses to acquire his or her first language (Johnson, 2004, 30-33). His viewpoint has been criticized on several fronts. The unfolding of the UG through maturation of the child is unknown; furthermore, it is unclear where, if anywhere, the UG exists since Chomsky defined it as a “mental organ … [lacking] a clear demarcation line between physical organs, perceptual and motor systems, and cognitive faculties” (1980; in Johnson, ibid; 33). Still, as a cognitive theory, the UG eliminates the need for behaviorist environmental factors, which are, from a cognitive linguistic standpoint, considered of less import than cognitive processes aiding the processing of linguistic data (ibid, 34).
In addition to the UG, Chomsky considers that language use is more than grammatical competence: pragmatic competence is required (ibid). Competence, according to Brown (2007a; 35) is the knowledge of the underlying rules of a system, event or fact—the ability to do something, though not necessarily the realization of the ability. This contrasts with performance which is the actual doing of an act. In language, pragmatic competence—linguistic competence in Bachman’s terminology (1990; in Skehan, 2007; 157-161)—is the knowledge of the rules and systems of grammar (Johnson, ibid; 87). Based on Ferdinand de Saussure’s (1916) concepts of parole (performance) and langue (competence), Chomsky’s view of language performance reminds us of an ideal language produced by an ideal user in ideal circumstances, since actual performance of language is “fairly degenerate in quality” given the errors inherent in actual situational usage (Chomsky, 1965).
Chomsky’s view about language performance quality, noted above, answer questions about why first language speakers may fail to produce utterances of accuracy in their first language, but it fails to answer why second language learners often fail to acquire their second language. Stephen Krashen’s input hypothesis model takes this into consideration. For Krashen, second language learners have access to the LAD, but access is mediated by an affective filter, “a metal block that prevents acquirers from fully utilizing the comprehensible input they receive for language acquisition” (Krashen, 1985: 3; in Johnson, 2004: 48). Comprehensible input is defined by Krashen as input which is just above the level of the learner—i + 1 (Brown, 2007a: 295). A reduced or eliminated affective filter leaves the LAD free to process input and L2 acquisition will, according to Krashen, occur.
Krashen’s theories are highly contested: the affective filter’s operation, as defined by Krashen, appears to represent the result and not the cause of the block to the LAD: it is the learner who is responsible for the affective filter’s operation (ibid, 49). Second, since it relies on Chomsky’s UG which allows for natural growth of language through an environmental “trigger”—a passive response to language growth— Krashen’s model has difficulty reconciling the active nature of input, which proceeds from cause to effect in language learning: input causes acquisition (ibid, 48). Third, it is difficult to understand what Krashen means by comprehensible input which is just above the learner’s ability. Brown notes its approximation with Vygotsky’s Zone of Proximal development, but that other linguists have pointed out Krashen’s simple formula makes it appear as if we are able to define “i” and “1” yet we are unable to do so (Brown, ibid: 297; see also Gregg, 1984; White, 1987).
Other cognitive linguists have modified Krashen’s model. Swain, for instance, suggests that comprehensible output is important to generate comprehensible input (Johnson, 2004: 51-52; see also Skehan, 2007: 16). For Swain, it is the noticing the disparity between “what I can say” and “what I want to say” that produces acquisition in an L2. Swain’s model identifies three main functions: noticing/ triggering or consciousness raising, hypothesis-testing, and reflection and places much of the process in the learner’s sensitivity to the difference between intended and apparent meaning of output based upon the resultant input. Output which presents input different than that expected results in a modification of the approach to output. Swain’s model is similar to that of Long who ostensibly considered negotiation for meaning as crucial for language acquisition, including the implicit rather than explicit negative feedback possible in communication.
For Long, the primary difference is in the interplay between the learning environment and the learner via negotiation of meaning. Negotiation for meaning is a combination of environmental, learner selective attention and L2 processing factors leading to language acquisition. Attention is crucial, implying the learner is able to identify linguistic features for change within his or her purview. Long, Johnson notes, does not advocate the explicit teaching of grammar, a focus on forms, each item sequentially and additively presented. Instead, Long and Robinson discuss a focus on form where attentional resources are allocated towards meaning through the use of grammatical forms (ibid, 55).
2.5 Sociocultural Theory as a Locus of SLA
In contrast to both the behavioral and cognitive schools of thought, a constructivist viewpoint sees learning, including language learning, as a socially constructed activity. Language is not, it may be noted, produced for its own sake. Language production takes place in a social context where meaning is jointly produced. Acquisition of language is likewise socially constructed and its meaning is likewise discovered (Brown, 2007a: 12-13). In the constructivist camp, Lev Vygotsky’s concept the Zone of Proximal Development (ZPD), which has been applied to language teaching, appears to be critical for an effective tool for teaching and learning. The ZPD functions in a social context, setting up roles for both mentor and learner; it is
the distance between the actual developmental level as determined by independent problem solving and the level of potential development as determined through problem solving under adult guidance or in collaboration with more capable peers.
(Vygotsky, 1978: 86)
Development is not an entirely solipsistic activity. Periods of solitude and of interaction with peers of higher development are both necessary to effect growth. Interaction in the sense identified by Vygotsky is between a novice and another who has mastered a particular skill or task. The concept of the ZPD has been formalized into a series of six protocols for scaffolding learners to mastery of a task:
- recruit interest in the task,
- simplifying the task,
- maintaining pursuit of the goal,
- marking critical features and discrepancies between what has been produced and the ideal solution,
- controlling frustration during problem solving, and
- demonstrating an idealized version of the act to be performed.
(Donato, 1998, 41; emphasis in Johnson, 2004: 130-131)
Each of the stages of scaffolding are the kinds of acts which a more capable peer does to foster learning in an individual. Donato, however, identified not only that scaffolding of this sort assists the learners, but that learners of the same level as well working individually can take advantage of the same techniques through self-talk (ibid; see also Lightbown and Spada, 2009: 48). In the use of drama for language learning, students are often given tasks for which they must figure out the best means of expression both individually as well as in peer groups as well as receiving direct scaffolding of the sort identified by Donato in the list of protocols, above. One exception, however, is the use of the word, idealized, in the final item. This will be taken up in the discussion about drama in the fourth chapter.
*(More needed)
Chapter 3 The Physical Basis of Language and Narrative Construction
The previous section dealt with some of the historical principles undergirding our research. In this chapter, we shall consider some more recent research into language which does not rely upon a purely behavioral or cognitive view of language acquisition. This chapter will discuss the physical embodiment of language particularly in reference to metaphorical usage. In addition, we will discuss the narrative construction with an eye towards its use in the formation of identity.
3.1 The Physical Embodiment of Language: The Neural Theory of Language (NTL)
That language, and especially metaphor, is considered by many to be embodied physically is not new, yet it is arguable. Language in Saussurean structuralism is the product of socio-psychological phenomenon (Taylor, 1992: 78)—reminiscent of a Vygotskian perspective rather than a Chomskeyan one. For Saussure, a speech community shares a common set of signs—the signifier and the signified—or language; membership means internalizing the signs of the community (ibid). Language is a vehicle of subjective expression and intersubjective understanding (80). For Chomsky, who agreed with Saussure’s concept of langue and parole, language has a natural genesis; every child born has access to all language grammars, but given time refines his or her language capacities to a single one—the child’s first language. Langue is outside the control of the will of the individual, imposed by outside members of the speech community: for communication to occur, two people must agree as to the meaning of a sign and its significance; what is taken to be a sign and what its significance are arbitrary, yet decided upon from without (82-84).
It is becoming more difficult to rule language a purely mental set of arbitrary constructs designed to allow communication of abstract concepts. The physical embodiment of language is so compelling an idea, that effort has been spent to identify a genetic basis for language, the FOXP2 gene, ubiquitous in the human population (Lai, Fisher, Hurst, Vargha-Khadem, and Monaco, 2001). Subjects having an impaired version of this gene appear to have severe linguistic impairments ranging from articulation to comprehension and grammatical difficulties (Pinker and Jackendoff, 2005; 218). Others look not to genetics but to a neurological basis for language perception and use.
Lakeoff and Johnson reject a purely philosophical view of language for one established in the neural network of the mind and body, a sort of “hidden hand” which shapes how experience is to be conceptualized (1999: 12-13). Considering a neurobiological perspective, Liberman and Mattingly (1985) claim in their motor theory of speech perception that objects of speech perception are “intended phonetic gestures of the speaker, represented in the brain as invariant motor commands that call for movements of the articulators through certain linguistically significant configurations” (2), the idea being that the gesture and the word are one and the same and are intimately linked in such a way that a translation from gesture to speech is automatic (ibid, 3)—they are “different sides of the same coin” (ibid, 30).
Liberman and Mattingly take a pragmatic view of speech and listening structures of cognition, arguing the difficulty existing for how two structures processing the same information could come about (ibid). Lakeoff and Johnson (1999) take up this position, proposing the efficiency of neural structures to both perceive input of a particular kind and to create motor schema for acting upon that input:
The visual systems of our brains are used in characterizing spatial-relations concepts. Our actual motor schemas and motor synergies are involved in what verbs of motor movement mean. And the general form of motor control gives general form to all our actions and the events we perceive. The point is this: In such models, there is no absolute perceptual/ conceptual distinction, that is, the conceptual system makes use of important parts of sensorimotor system that impose crucial conceptual structure.
(ibid, 39)
The physical reality of the body’s perception system actually creates the cognitive systems whereby information is perceived. In other words, what the eye sees and the ear hears creates the systems which the eye and ear use to further perceive. This idea is illustrated in the metaphor Knowing is Seeing: Christopher Johnson (1997, ibid, 86) notes that acquisition of conceptual metaphors goes through two stages—conflation and differentiation—where the verb, see, used as a child might as in “See what I spilled?” does not mean the same as in “See what I mean?” In the first instance, see indicates a gaining of knowledge—to see and know—whereas the second it is a metaphorical usage meaning know even though no physical seeing has taken place (ibid). The second usage indicates a refinement upon the first and would not be expected in earlier language developmental stages.
To deepen our examination of the physical embodiment of language, we find many studies in which an approach or methodology to language learning which is physically involved or one which activates schematic metaphoric knowledge, outperform a more traditional approach or methodology. Lindstromberg and Boers (2005) showed in experiment that students within an enactment-based (ie: physical action) foreign study program utilizing James Asher’s TPR methodology outperformed control groups in terms of vocabulary recall and retention, particularly with manner-of-movement verbs like sway and hurl. This seems to confirm Asher’s own studies (1965, 1966 and 1969) that activating the motor system during language learning is an effective means to improve retention and recall. This may be due to the lateralization of information across brain hemispheres as the right hemisphere uses action for expression, bypassing the analytical left hemisphere (Mangubhai, 1991: 269). In experiments where learners were simply cognizant of the source domains from which metaphors might be categorized, there appear to be benefits from enhanced metaphor awareness (Boers, 2000: 562). Boers suggests that this may be due to benefits from image processing, employing cognitive effort to identify source domains and making categorizing judgements promoting deep-level cognitive processing; and metaphoric themes as categories may also provide a framework for lexical organization (ibid, 563).
Boers’ experiments employed categories similar to those identified by Lakeoff and Johnson, (2003), such as ANGER IS HEAT, BODY IS A CONTAINER FOR EMOTIONS, MORE IS UP; LESS IS DOWN; and so on (Boers, 2000). These categories are immediately comprehensible in terms of imagery and are both physically and conceptually salient. When speaking of metaphors, Lakeoff and Johnson intend metaphorical concept (2003, 6); furthermore, concepts become so because of their inferential capacity,
their ability to be bound together in ways that yield inferences. An embodied concept is a neural structure that is actually part of, or makes use of the sensorimotor system of our brains. Much of conceptual inference is, therefore, sensorimotor inference.
(emphasis in original; ibid, 20)
A theory of the physical embodiment of language rests upon the neural theory of language (NTL) proposed by cognitive science in which a gestalt position is taken: both motor schemas and functions a thing can perform are related by the part-whole structure. Mind is not independent of the structure wherein it exists; things with bodies can only perceive and perform within the limits of the body; perception is based upon basic-level categories such as cat, mat, boy, and ball as well as swimming, walking and grasping rather than Tabby, yoga mat, lad, basketball, breast stroke, saunter or handshake. Moreover, extensions of the body extend our perception of the world at large—microscopes and telescopes, for instance allowed us to think biologically or universally, before which our concepts of the world were spiritual or limited to what we could see with the naked eye (Lakeoff and Johnson, 1999: 28-29; 37).
Lakeoff and Johnson, in proposing the hypothesis of the embodied mind, are clear to identify the foundation of this theory is based upon existence proof. They propose no neurophysiological proof such as PET or MRI scans of the structures used in perception and conception. Instead, they utilize three models of Terry Regier (1996), David Baily (1997) and Srini Narayanan (1997) (ibid: 40-41). Regier’s model was capable of using visual information to build, recognize and later make linguistic descriptions of a static geometric configurations such as “on” and dynamic configurations such as “onto” using world languages. Baily’s model identifies and names hand motions in the world’s languages, but also recognizes differences between languages to categorize hand actions. The model can both perform the correct function given the word, or given the visual stimuli, give the correct verb to describe the observation. From this, Bailey does not imply that the meaning of the verb “grasp”, for example, is the same thing as the physical action of the verb “grasp”; instead, the model implies the motor schemas and parameter values of the brain which observe and categorize can operate without physical response (ibid, 580). From this, it is reasoned, motor schemas that can and do control the body could function in reasoning when not controlling the body.
Narayanan’s model really describes the full program of a motor schema:
- Getting into the state of readiness
- The initial state
- The starting process
- The main process (either instantaneous or prolonged)
- An option to stop
- An option to resume
- An option to iterate or continue the main process
- A check to see if a goal has been met
- The finishing process
- The final state
(ibid, 41)
Narayanan’s model is capable of characterizing aspect—a crucial aspect of linguistic descriptions of events. Narayanan’s model shows that the same neural structure capable of performing motor control is likewise capable of controlling logical inferences about the structure of actions. His model was capable of performing metaphoric projections in an abstract domain—international economics—when given conceptual metaphors from news reports, such as “India loosened its stranglehold on business,” and “France fell into a recession and Germany pulled it out”. Taking these models as the foundation, Lakeoff and Johnson hypothesize that since brains tend to optimize and only add what is necessary, the sensorimotor system is utilized for both perception on a physical and an abstract (metaphorical) level rather than developing a completely new system (43).
3.2 Narrative and Identity Construction through Narrative
Significant to this paper is the use of narrative as a mode of discourse between participants. The ability to tell a good story as a highly regarded talent (McCarthy, 1991: 137) is taken as the critical teaching point of this study. In this study, participants create spoken and written narrative texts for analysis. William Labov presents a general framework for narrative construction. Narrative is defined as “one method of recapitulating past experience by matching a verbal sequence of clauses to the sequence of events which (it is inferred) actually occurred” (in Jaworski, 2006: 218). Narratives of personal experience are narratives which have entered into the biography of the speaker (Labov, 1997). Narrative structure is composed of six parts:
- Abstract
- Orientation
- Complicating action
- Evaluation
- Result or resolution
- Coda
(Jaworski, 219)
The abstract is a summary of the events which will follow—the reason why the listener or reader should pay attention to the text which follows. The orientation provides the who, what, when, and where of the narrative. The complicating action answers the question, “Then what happened?” (222), and identifies the chain of actions which constitute the main body of the text. All sequential clauses, Labov defines (1997), are clauses of complicating action and vice versa. Once narrated, the actions are evaluated, or validated (see Swan in McCarthy, 1991: 138), in terms of the participants—the “So what?” or the raison d’être of the text. This is the point of telling the story. Evaluations of the events happen by making a comparison between what happened and some alternative reality that was not realized; furthermore, with an increase in age, Labov notes evaluative statements occur with greater frequency (see Labov, 1972, in Labov, 1997).
The result or resolution tells what finally happened (224), after which there are no more complicating actions. Finally, narratives often have a coda which tells the listener or reader the narrative has come to a conclusion. Codas are often “free clauses” or generalizations showing effects of the events of the narrative (221); these free clauses refer to conditions which hold true during the entire narrative (Labov, 1997).
Labov is careful to identify that the above mentioned elements of narrative are “typical”, only the complicating action is necessary for a text to be a narrative (Jaworski, ibid: 221). With each complicating action clause, Labov notes, the speaker or writer evaluates the overall effect of the narrative, making choices as to whether the story is in a state of completion or not and providing evaluation when necessary (see figure 1, below).
Figure 1: Narrative Structure (Reproduced from Jawarski, 2006: 225)
Later, however, he defines a minimum narrative as a series of linked clauses containing at least once temporal juncture where (Labov, ibid). In addition, narratives are sets of bound, restricted and free clauses. Bound clauses are sequential, independent clauses. By bound, Labov intends temporal binding, and identifies the use of the simple past, the past progressive, the present tense with the semantic interpretation of a preterit as devices which, at least in English, which are necessary for narratives (ibid).
Stories are not told in a vacuum, but collaboratively (McCarthy, 1991: 139). Narratives are frequently considered to be composed of a single set of uninterrupted clauses. Sacks, as Labov points out, disagrees with this view. Conversations are composed of strings of at least two turns (Sacks, 1967; in Couthard, 1985: 69). Narrative is not defined by “holding the floor” but by controlling the assignment of the speaker, and this necessitates the interspersion of back channel signals. These signals are taken as turns of talk (ibid; see also Edwards and Middleton, 1986) and set up transition relevance and expectation (Sacks, 1967). These signals indicate the evaluation by the addressee as permission to continue speaking. These are summarized by Labov as follows:
- Successful completion of the narrative requires automatic re-assignment of speaker role after the following turn if the narrative is not completed in that turn.
- A narrative of personal experience must contain at least one reportable event.
- A most reportable event is the event that is less common than any other in the narrative and has the greatest effect upon the needs and desires of the participants in the narrative.
- The more reportable the most reportable event of a narrative, the greater justification for the automatic reassignment of speaker role to the narrator.
(ibid)
A paradox exists, in Labov’s evaluation: the more reportable the event the less credible is the narration of the event. Credibility is further related to causality. When the answer to the question, “How did that happen?” is appropriate to the telling of the story, the question is not and the clause in question is assumed to keep the cause-result structure. Moreover, while a narrator may keep the floor and tell a credible story, the teller expects agreement on the part of the listener, yet participants may disagree about the significance a particular story holds (Ryave, 1978; in Coulthard, 1985: 84).
Labov (1997) identifies in his examination of narrative a series of elements which seek to contain narrative within a single matrix of coherence. Among these are temporal relationships between clauses indicated through verbs in the simple past, past progressive, indicative mood, the use of subordinate conjunctions, and clause types belonging to the categories noted above (abstract, orientation, etc.). Placement within the narrative determines much of the identity and function of a particular clause as, for example, “evaluation is characteristically concentrated in an evaluation section”; this does not preclude the use of evaluation within any other part of the narrative, but identifies a general maxim for narrative structure. Sacks (1972; in Coulthard, 1985: 85) identifies ways in which any two clause items may be “heard” as belonging together. The first is a membership categorization device handling a descriptive category such as gender to which male and female, subordinate to the first category, both belong. In the clause pair,
The baby cried. The mommy picked it up.
both baby and mommy are understood to belong to the same device—family. Some categories will belong to more than one device: baby, for instance can belong to family as well as to stage of life. Sacks’ economy rule (ibid, 86) recognizes that utilizing a single category for any device is tantamount to paying adequate reference thus providing coherence; keeping consistency occurs when categories of the same collection of one device are used to categorize further members of a population. Keeping the consistency and the economy rules in the example,
The first baseman looked round, the third baseman scratched himself.
the basemen are heard as belonging to the same team. Sacks furthermore shows that activities are appropriate to members of certain devices—they are category-bound activities. Thus, as Coulthard states, crying is a category-bound activity for the category baby when it is a member of the stage of life device, but not necessarily for another category such as adolescent or adult.
Schlegoff (1972) outlines the concept of membershipping in which listeners must be brought into a particular topic by the use of particular information assumed to be known by the listener. In one sequence, Coulthard quotes Sacks’ (1968) example of an exchange on an airplane:
Passenger: Do you have a cigarette?
Stewardess: No we don’t. They don’t provide that service anymore.
whereby the stewardess is membershipped as a stewardess and the question is addressed and received in that role. Counter to this example, the stewardess may have considered to have been membershipped as a stranger and offered one herself. In avoid situations like,
A: I just came back from Irzuapa.
B: Where’s that?
Coulthard suggests membershipping listeners with a pre-sequence:
A: D’you know where the Triboro Bridge is?
B: Yeah.
A: Well, you make a right there.
Narrative may be understood as organized under a series of principles outlined by Grice (1975). These are highlighted in the following statement:
make your conversational contribution such as is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged.
(in Coulthard, 31)
Coulthard expands upon this principle by identifying four major areas in which speakers must keep their discourse organized:
-
|
relation |
be relevant |
-
|
quality |
a) do not say what you believe to be falseb) do not say that for which you lack adequate evidence |
-
|
quantity |
a) make your contribution as informative as is required (for the current purposes of the exchange)b) do not make your contribution more informative than is required |
-
|
manner |
a) avoid obscurity of expressionb) avoid ambiguity
c) be brief
d) be orderly |
These maxims are not rules set in stone, as Coulthard notes; rather, they are suggestions for constructing discourse. In terms of narrative, these same rules can be seen to apply since, first, narratives which are not relevant to the discourse topic at hand (ie: no membershipping, coherence, topical consistency or category/ device consistency) will be considered inappropriate to share; second, stories which are judged to be fabrications quickly destroy the credibility of the speaker; third, a speaker whose narrative is overwrought with detail will eventually result in increased negative backchannel signals; and fourth, narratives which fail to be temporally coherent, short or unambiguous will likewise receive negative backchannel signals. Narratives which fail Grice’s maxims will lack reportability, to use Labov’s term.
Identity formation through narrative is about how the teller of the story decides to reveal his or her personal character through topic choice, interaction, description, analogy, referencing, inferencing, mood, grammatical choice or evaluation of the events. These are reflected in the evaluation of narrative, as outlined in Labov (1997; see also Jaworski, 2006). Labov notes that it is in the sentence grammar which gives direct clues to evaluation. For example, the speech of an actor in a narrative animated to speak directly (at the level of the narrative—ie: reported speech of the event) is open to evaluation no matter the topic or evaluation. Negatives, comparatives, modals or futures can be seen as a form of evaluation. In the example statements,
I don’t even wanna fool with ya.
and
An’ a guy told me, says, “Don’t move your head.”
the first statement can be compared to a situation in which the speaker would want to fool with the other narrated character; the second can be compared to a situation in which it would be safe for the speaker to move his head. (More to come…Kiernan Dissertation, part 3; Section 4.1, pg. 157)
A complication for L2 learners is the fact that, for lower-level learners, moment-by-moment lexico-grammatical encoding at the clause level tends to interfere with the discourse-level skills resulting in a skeletal narrative of bare facts with little evaluation (McCarthy, 1991: 140). In addition, typical published activities tend to favor single instead of joint storytelling.
3.3 Corpus Linguistic Analysis of Multimodal Discourse
Language has long been known to be composed of rule-governed patterns into which certain words were understood to fit to the exclusion of others. Typically, these rules are the grammar which organizes language, yet within recent decades, collocation has been recognized as a significant organizing system operating within the grammar. J.R. Firth is among the first to identify collocation with his statement, “you shall judge a word by the company it keeps” (Partington, 25). From this humble beginning, corpus linguistics has grown.
John Sinclair identified two properties of language showing that textual coherence does not happen at random—the idiom principle and the open-choice principle are significant to collocation. Should a word fit within a set phrase, then the idiom principle is at play; conversely, if there is no set phrase, then the open-choice principle is in use (Hunston, 124). For example, the words in the phrase, of course, as an expression of assent, have meaning together; yet when another word is substituted—by course, the phrase loses its meaning. The open-choice principle is the slot-and-filler model of language description where a large series of complex choices is made to produce coherent text (Sinclair, 108-109), the choice proceeding limiting those following. These two principles provide the foundation for collocation.
Textual meaning accumulates globally through many local collocations. Language description by collocation alone would be meaningless without some way to show relevance. Sinclair identified that a native-language user’s intuitions about language, will be unrevealing in terms of usage (Sinclair, 39). Having amassed a great amount of language knowledge, most personal evaluations will be “ideas about language rather than facts of it.” Corpus linguistics combines text with technology to reveal language facts. The introduction of digital computers and concordancing software from the mid 1970’s onwards provided the muscle with which to do most of the work of finding collocations (Kennedy, 5-7). Electronic databases held the texts in a machine-readable format. Software could identify collocations as well as provide linguists with statistical data showing which were most relevant. Knowing how basic concordancing software works can provide useful information on the evidence returned from a search (McHardy, 15). Collocations with higher values of frequency, t-score and mutual attraction are more statistically relevant (see Table 2).
Table 2 Collocation Examples Ordered by Frequency and T-Score
Collocate |
|
Frequency |
|
T-Score |
his |
|
847 |
|
25.051640 |
her |
|
661 |
|
23.187461 |
at |
|
435 |
|
14.081992 |
under |
|
154 |
|
11.160451 |
she |
|
238 |
|
10.843065 |
|
|
|
|
|
(Hunston, 14-16) |
A limiting factor in the importance of these measures, however, is the amount of text under examination. A smaller text will produce higher values, thus inflating the importance of the collocation. This is particularly important when dealing with texts of different sizes, as a larger text will show a particular feature to me less statistically relevant while a smaller text will show the reverse.
Corpus discoveries, regardless of size, are also limited by the way in which the sorting software is set (McHardy, 15). Setting the software to find a single word—hard, for example, will return collocations for the word hard, but not for other forms such as harder, hardest, or hardly. Searching for the lemma hard@ will return forms of the same type. Searching for hard*, however, will provide all the words which have hard as a part as well as all other forms besides. Searching for a lemma instead of the single word may appear to bear more fruit, yet the time and energy required for sorting is greatly increased.
Table 3 Lemmas of Hard, Difficult and Challenging
|
hard |
difficult |
challenging |
-
|
hard |
difficult |
challenging |
-
|
harder |
|
challenges |
-
|
hardest |
|
challenge |
-
|
|
|
challenged |
|
A lemma is particularly useful for verbs, but at times may include both verb and noun forms, as in contact, which, in both forms, includes contact and contacts. Collocations alone cannot determine when one or the other form is in use while both may contribute to a word’s frequency, t-score or mi-score. Concordance lines may highlight such use (see Table 4), as will setting the software to discover noun or verb forms with “/NOUN” or “/VERB” (among others).
Table 4 Concordance Lines for the Lemma Challenging
Authors of a most informative and |
challenging |
book. A Nation in Denial; The |
who is challenging Bush and |
challenging |
Bush’s policies. Instead, we’re |
differences now coming out and |
challenging |
dearly-held values of community |
The idea of upping the ante and |
challenging |
Governor Clinton to more |
While this is a very difficult and |
challenging |
time, the board is confident |
|
Lexical patterning, as indicated by concordance lines is essential for discovering elements of language which are both statistically relevant and which provide teachable systems which are broad enough for use between registers and genres. The above lines of concordance indicate some patterns which challenging follows (see Table 5).
Table 5 Some Patterns of Challenging
- very ADJ and challenging NOUN
- 2. most ADJ and challenging NOUN
- 3. challenging NOUN
- NOUN and Challenging NOUN
|
The items in Table 5 are at the same time communicable, teachable and general enough to cover different registers and genres. Furthermore, they can be used to check for similarity between other lexical items. Hunston and Francis (1998) note both that lexical items have patterns and that similar items fall into groups based on shared aspects of meaning. In the first pattern displayed in Table 5, the word hard could be substituted for difficult in the adjective position, but good could not without changing the meaning of the passage. Additionally, each of these patterns will collocate in statistically different ways; however in terms of a particular lemma, one form will typically be more common than another (Stubbs, 172).
Corpus study reveals features as synonymy and semantic prosody as important features in language. Synonymy is useful for language learners since learning which words are similar to others creates meaningful links between vocabulary items, making them easier to acquire (Partington, 39). In addition, synonyms reduce repetition of words within text as well as allow users to make semantically similar statements across different registers. Synonymy refers to having the same sense, though not the same reference. Lyons (1981, in Partington, 40) distinguishes between complete synonymy and absolute synonymy. Items are completely synonymous if they are descriptively, expressively and socially identical in meaning and absolutely synonymous if they have, in addition, the same distribution. Furthermore, lexical items are descriptively synonymous if, when they are interchanged, the message in which they appear is not affected. Partington notes that “the selection of one rather than the other may change the social or expressive meaning of the utterance, but hold constant its descriptive meaning (if it has descriptive meaning) in which case, we can say that the intersubstitutable lexemes are descriptively synonymous” (40).
Semantic prosody was first described by Sinclair (1991) as meaning extending over more than one unit, shown by Sinclair with the phrase, set in, in which he shows that the phrase, while not inherently negative, collocates with items such as rot, decay, disillusion, and infection and points to a negative state of affairs extending over set in (Stubbs, 173). As “a subtle element of attitudinal, often pragmatic meaning” (Sinclair 1998: 20), semantic prosody determines the meaning of the whole unit (Mahlberg, 33). Both synonymy and semantic prosody can provide language users and learners with relevant information about when a language community utilizes which items for purposes which are culturally relevant such as the connotative value of a unit of language.
Observing the lines of concordance, above, we can see that the closest words to challenging are the grammatical words and and the, the nouns time, Bush, Governor Clinton and values, and the adjectives difficult, informative, and dearly–held. We might conclude from this brief observation that challenging is most often used as a verb with object nouns—in particular political figures and policies. In fact, a wider search reveals many other categories to which challenging belongs. Even more relevant is the idea that challenging authority is a cultural activity which is valid in certain circumstances. Whatever may be said of a culture based on its language, knowing when a challenge to authority is vitally important which may allow one to more fully participate in a foreign culture.
While corpus information serves a critical function in language instruction and may show which patterns are most prevalent within a particular genre or register of language (Altenberg, 14), it can show which lexical items collocate within those patterns. Furthermore, it can show users ways in which cultural ideas are conveyed in language. Kachru (1994) argues that users of a language exist in three concentric circles, with native users at the center and non-native and learners further out. These circles coincide not only with language ability but also with social inclusion/ exclusion. Such division, considered offensive by some researchers, is often exacerbated by the fact that language learners are often instructed by non-native speakers (Hinkel, 928), further distancing the learner from the centre. Inclusion closer to the center of a group of language users may be encouraged by cultural information imbibed from corpus inquiries. Corpus information can assist in making the divisions between native and non-native speakers less distinct, allowing learners greater social and political access to the world of the studied language.
Corpus linguistics affords great advantages in the searching of text for patterns of behavior, yet, recognizing that discourse—particularly narrative for the purposes of this study—does not exist in a vacuum, a multimodal corpus can provide a richer consideration of L2 learner language usage and L2 identity formation. The term multimodal refers to the usage of multiple types of representation in order to convey meaning. Gunter Kress identifies script, image, and music as some of our oldest modes of communication. These he calls modalities and adds to them others such as speech, gesture, framing, dance, design, pace, and knowledge production, noting that the use of one over another modality has implications for knowledge and identity (Kress, 2010: 12). Framing refers to the rhetorical gestures (ie: metaphor) one makes when constructing a communiqué in any particular modality.
Kress notes that there is never a single mode which works in isolation; in communication, several modes are always used together, “in modal ensembles, designed so that each mode has a specific task and function” (ibid, 28). These ensembles are designed—that is, resources are arranged and selected—to create a particular meaning using a particular rhetorical analyses—aims and purposes of the rhetor—instantiated through choice. As noted above, it is the concept of choice which is most salient when considering the evaluative mode of narrative; it is choice which marks the identity of the narrator whether intended or not. Following a social-semiotic theory of multimodality, laid out by Kress, narratives are compositions of signs,
always newly made in social interactional signs are motivated, not arbitrary relations of meaning and form; the motivated relation of a form and a meaning is based on and arises out of the interest of makers of signs; the forms/ signifiers which are used in the making of signs are made in social interaction and become part of the semiotic resources of a culture. The relation of form and meaning is one of aptness, of a ‘best fit’, where the form of the signifier suggests itself as ready-shaped to be the expression of the meaning—the signified—which is to be realized. Aptness means that the form has the requisite features to be the carrier of the meaning.
***