Having a say Direct speech representation in Greek youth storytelling

1. Introduction Speech representation, namely the report of words that have been uttered in a spatiotemporal context other than that of the current interaction, is a phenomenon that has been examined in various discourse genres. Different approaches have been adopted relating to the forms and functions it embeds and, depending on the approach, various terms have been employed to describe itsmeaning, such as speech report, reported discourse and speech presentation.1 Traditionally referred to as reported speech, it has been examined both in written and spoken language, ranging from traditional grammar studies to text-oriented analyses. The latter includes among others literary narrative, media texts, academic writing and spoken language (see among others Voloshinov, 1973; Coulmas, 1986; Clark and Gerrig, 1990; Mayes, 1990; Chafe, 1994; Myers, 1999; Holt, 2000; De Fina, 2003; Semino and Short, 2004). Journal of Pragmatics 43 (2011) 3374–3386 Gender identities Socio-cultural norms powerful narrative device which contributes to the foregrounding of gender identities and helps sustain social stereotypes. Finally, I observe that the identity repertoire available to speakers is socioculturally determined. � 2011 Elsevier B.V. All rights reserved. Contents lists available at ScienceDirect Journal of Pragmatics journal homepage: www.elsev ier .com/ locate /pragma This study explores direct speech representation (henceforth DS) in Greek adolescents’ conversational narratives. In particular, I investigate how the use of direct speech contributes to bringing forth emerging gender identities. Following studies that have shown how DS plays an important role in presenting oneself and/or others, I address the aspects of variation of DS (see among others Hamilton, 1998; De Fina, 2003). For this purpose, my analysis focuses not only on the different functions performed through this device, such as vividness and dramatisation, but also on different elements of variation within direct speech, for example the various voices represented through DS, the status and gender of the represented participants (speech producers and addressees), as well as the kinds of represented acts. This type of analysis Having a say: Direct speech representation in Greek youth storytelling Sofia Lampropoulou * Department of Linguistics and English Language, Lancaster University, Lancaster, LA1 4YT, UK A R T I C L E I N F O Article history: Received 25 October 2009 Received in revised form 12 July 2011 Accepted 18 July 2011 Available online 8 September 2011 Keywords: Speech representation Storytelling Direct speech A B S T R A C T This paper explores speech representation in Greek youth storytelling. In line with approaches that view reported speech as a means of self and other presentation, it investigates how direct speech contributes to bringing forth speakers’ identities. I adopt a quantitative as well a qualitative analysis to address direct speech in terms of different elements of variation, namely speech producers and addressees, gender and status of speech producers and addressees, represented acts and represented interactions. Drawing from the dynamic, constructed nature of speech representation, I argue that the reporting choices of the youths in my data reflect the norms of masculinities and femininities which are in social circulation in their everyday worlds. Direct speech is therefore viewed as a * Tel.: +44 7788895764. E-mail address: s.lampropoulou@lancaster.ac.uk. 1 The terms ‘report’, ‘representation’ and ‘presentation’ have often been used interchangeably to describe ‘speech’ or ‘discourse‘. I will be using the term ‘representation’ because it implies (a) that there is an anterior/original discourse situation, which is reported/represented in the posterior discourse and (b) that the relation between, for example, direct speech and the speech in the reported situation is not ‘unproblematic’ (see Short et al.: 336). Both the above are of crucial significance to my data analysis. For a detailed discussion on the terminology, see Short et al. (2002) and Lampropoulou (2007). 0378-2166/$ – see front matter � 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.pragma.2011.07.009 employs quantitative tools which correlate the aspects of variation mentioned above. Drawing from quantitative results, I will show that my informants opt for different reporting choices in the stories they tell in order to position themselves in differentways. Additionally, I will provide two representative narrative extractswhich highlightmy quantitative results and point to the emergence of gender identities through the use of DS. Within this approach, DS is viewed in its entirety, as it is examined through the different forms it takes, the different aspects of variation it involves and the various functions it performs. S. Lampropoulou / Journal of Pragmatics 43 (2011) 3374–3386 3375 2. Direct speech representation in spoken language The observation that language is inherently contextual has its roots in the philosophy of language, where Bakhtin (1986: 91) remarks that ‘‘any concrete utterance is a link in the chain of speech communication of a particular sphere [. . .]. Every utterancemust be regarded primarily as a response to preceding utterances of the given sphere’’. When people use language they rely on some mutually understood relationships that may relate to the context in which an utterance occurred. Every utterance refers back to others and cannot be separated from its surroundings. In other words, an utterance cannot stand on its ownwithout taking into consideration previous or future utterances, as ‘‘words bring with them the contexts where they have lived’’ (Bakhtin, 1981: 293). Within the same concept, research on speech representation in spoken language has argued that direct quotations in spoken contexts must count as representations or reconstructions of speech and not as actual reproductions of the original utterances2 (see among others Tannen, 1989; Baynham, 1996; Holt, 2000). This is due tomemory limitations3 but also to the fact that when utterances change context, there is an inevitable interaction between the current reporting context and the representedwords (see also Dubois, 1989; Sternberg, 1982). Thus, ‘‘one cannot speak another’s words and have them remain primarily the other’s words’’ (Tannen, 1989: 100). Speech representation is viewed as a dynamic process, as it inevitably carries the subjective contribution of the current speaker. The use of direct speech representation in everyday conversation can be further explored in relation to Erving Goffman’s (1981) work on the three roles of ‘speaker’, namely ‘animator’, ‘author’ and ‘principal’. The ‘animator’ is responsible for uttering the sounds that result in intelligible speech. The ‘author’ is the one who selects ‘‘the sentiments that are being expressed and the words in which they are encoded’’ (Goffman, 1981: 144). Finally, the ‘principal’ is the person whose beliefs, values and attitudes are being expressed. In some cases, the three speaker roles can merge into one, but often they do not. Considering the three subsumed roles of speaker, Goffman (1981: 144) observes that when someone reports the speech of another person (s)he is the animator but not the author or principal. Therefore, he concludes, the ability of a speaker to switch roles throughout the interaction enables him/her to use a specific type of language, such as taboo language, with greater freedom. This means, for example, that a speaker might decide to swear when representing the words of another person (therefore acting as animator) rather than when speaking in his/her own voice (therefore acting both as animator and author). As Goffman (1974/1986) argues, a speaker who uses taboo language when reporting the words of another person intends to "stand in a relation of reduced personal responsibility for what he is saying. He splits himself off from the content of the words by expressing that their speaker is not he himself’’ (Goffman, 1974/1986: 512). In this paper, I will present that my narrators, as animators – in Goffman’s sense – of their stories are able to construct the narrated events, including the represented words. Put differently, they are able to re-shape the original utterances and, at the same time, to take a stance on the represented words. This may involve the exclusion and/or inclusion of represented voices, the highlighting and/or downplaying of represented words and the detachment from or involvement with the original words. Constructive strategies like this confirm Bakhtin’s ideas that speech representation is an active process of transformation. The essential point is that one does not represent another voice unless it serves a purpose in the current context (Baynham and Slembrouck, 1999: 451). I will argue that, in the spoken data under analysis, the different functions of DS have to be addressed in parallel with the reporters’ communicative purposes, namely to present themselves and other characters in a range of ways, depending on the reporting context. In the following section, I focus on the use of DS in one of its most frequent environments of occurrence, that is, storytelling. 3. Direct speech in narratives One of the properties of direct speech is that it seems to reflect the perspective of the original speaker and not that of the reporter, because it maintains the person, space and time deixis of the original words. To this end, it contributes to the dramatisation and vividness of the words being reported and therefore, to a successful and effective story. Polanyi (1985) argues that storytelling is a social act, and as such, it is governed by social norms. One of these norms is the newsworthiness 2 Traditionally, direct forms of speech presentation have been associated with faithfulness, in the sense of verbatim reproduction of the original utterance. This is because in most forms of speech presentation the reported words (via the grammatical changes that take place) reflect the perspective of the reporter and the addressee cannot be certain about the words that were originally uttered. On the other hand, direct forms of speech (re)presentation involve a series of grammatical changes, such as the shift in person, time and space deixis, which reflect the perspective of the original speaker. Therefore direct forms seem to evoke the original voices and carry implicit assumptions of faithfulness. 3 Psychological studies have shown that verbatim recall is often impossible, because subjects tend to remember the meaning of the utterances rather than the actual form (see Anderson, 1974; Lehrer, 1989). and tellability of the story; this means that a story has to be interesting and worth telling, or else narrators are likely to be negatively judged by their audiences (Labov, 1972). Narrators aiming at the newsworthiness of their stories employ specific linguistic and non-linguistic features that contribute to effective storytelling, including speech representation. Within this concept, Mayes (1990: 350) observes ‘‘What better way to showwhat the point of a story is than to dramatise it and act it out using direct quotation?’’ Both the content of the represented words and the ways they are strategically positioned throughout the stories serve as indirect evaluative devices on the part of the narrators (see also Hill, 1995). This is mainly S. Lampropoulou / Journal of Pragmatics 43 (2011) 3374–33863376 achieved by allowing the audience to draw their own conclusions on the reported words which, displaced from a distant past, are brought to the forefront of the story. Based on the above, the analysis of DS in narrative is a powerful tool for understanding how narrators construct an effective and tellable story, serving their conversational goals (see among others Labov, 1972; Mayes, 1990; Georgakopoulou, 1997; De Fina, 2003; Clift and Holt, 2007). Recent research on discourse analysis and interactional sociolinguistics has put forward a dynamic approach to the analysis of identity within the framework of social constructionism (see among others Coupland, 2001; De Fina, 2003, 2006; Bucholtz and Hall, 2005; Georgakopoulou, 2005; De Fina et al., 2006; Benwell and Stokoe, 2006). Specifically, it is argued that identities are socially constructed and emerge discursively through situated discourse. This view challenges earlier approaches within the framework of essentialism which treat identity as a pre-given property residing in peoples’ minds. Identity is not viewed as essentially defined but as a social constructionwhich involvesmultiplemanifestations. To this end, it is argued that identity is not one stable and fixed entity but that different aspects of identities become salient in different contexts of communication. Within this approach, people employ different linguistic resources in order to negotiate, transmit and display different aspects of their identities (see De Fina, 2006). However, even within this dynamic approach, the social identity repertoire available to speakers cannot be inexhaustible but restricted, as it is socio-culturally determined. Thus, although constructionist theorists argue that people construct their identities in unpredictable ways, this process is always in accordance to social circumstances and interlocutors (see De Fina, 2006: 253). I argue that the kinds of identities my narrators reserve for themselves and other characters through the use of direct speech, are subject to the norms and restrictions of the broader socio-cultural environment they currently inhabit. 4. Data and the informants The data I am using constitute part of a broader set of narrative data that have been collected for the purposes of a longer research project in which I have previously participated as a researcher.4 The project involved the collection of naturally occurring conversational narratives among youths living in the city of Patras in Greece. For the purposes of data collection, a group of four researchers, including myself, visited two different schools in Patras and attended classes of the last grade of high school over a period of almost twomonths. Our attendance facilitated familiarisation with some of the students, as our main objective was to start meeting outside the school environment in order to eventually record casual interactions. After being acquainted with the students for almost two months, we recorded authentic conversations between male and female students that resulted from our interactions with them. Recording took place at school, in coffee shops, in the researcher’s car, at the informants’ homes and over the telephone. The informants felt comfortable talking with us since we had become members of their groups, following consistently all their in-school and outside school social activities. It should be noted that the very small age difference5 was one of the factors that facilitated involvement. Overall, the project includes 24 h of recorded interactions between adolescents and researchers. I focus on ten single-sex conversations recorded by myself, each lasting approximately an hour. Four of these consist of all-male informants and six of all-female informants. In most conversations the adolescents have been recorded in pairs whereas some include three or four informants. In total, there are 21 informants, as some of them participate in more than one conversation. All the informants are aged between 17 and 18 years-old and attend the last grade of high school. They have been born and raised in the city of Patras and they all live with their parents. They belong to a nuclear family, in the sense that their families consist of both a father and a mother and children who live together. In few cases, older siblings of the informants are university students studying away from Patras or have been married and hence they do not live with them. The adolescents’ everyday practices revolve around school attendance and activities and attendance of private lessons/tutorials. The latter prepares them for the exams students have to take in order to be able to study in higher education. Not all students share expectations to study in higher education. However, during the time the data was collected, it was common practice for all students to go through entry exams. Depending on their performance in the exams, they would consider their options for further studies. In the discussion section, I will employ the background information concerningmy informants in order to underpin my arguments relating to the emergence of different aspects of identities through DS. 4 Project title: ‘A sociolinguistic study of the narratives produced by youths in Patras, Greece’, K. Karatheodoris 2425, funded by the Research Committee of the University of Patras, Greece, Principal Investigator: Dr. Argiris Archakis. For further information about the project, see Lampropoulou (2007). 5 Studentswere 17–18 years old and researcherswere strategically selected to be in their twenties, i.e. in their second or third year of their undergraduate studies. The ten conversations under scrutiny include 230 spontaneously told narratives,6 in the sense that they were produced without the invitation of the researcher. As the specific focus of this paper is DS in storytelling, I isolated7 918 DS instances included in these narratives. These have been analysed in terms of various elements of variation which are presented below. 5 c. d. e. g. S. Lampropoulou / Journal of Pragmatics 43 (2011) 3374–3386 3377 �turn taking. 5.2. Represented act After considering the represented agents, that is both the represented speaker and addressee, I went on to consider the represented acts. More specifically, I examined the types of speech produced in the DS instances. For this purpose, I adopt the term rapport orientation (Spencer-Oatey, 2000: 29) which relates to the use of language in order to promote, maintain or threaten harmonious social relations between interlocutors and involves the following values: � Rapport enhancement orientation: a desire to strengthen or enhance harmonious relations between interlocutors. This would include for example the exchange of compliments and utterances that signal agreement with the addressee. 6 Narratives have been defined according to Labov’s (1972) criteria on narrative structure. 7 For the purposes of quantification, I established boundaries between different DS instances every time there was a change in the person who produced the direct speech and/or in the setting (time and/or space) where the represented utterance takes place. For further discussion, see Lampropoulou (2007). 8 For similar distinctions, see De Fina (2003) and Archakis and Papazachariou (2008). t cases where it cannot be identified who the addressee is). Represented interactions: In this case I deal with whether a DS instance forms part of a represented dialogue where turn aking takes place or whether it consists of a single reported utterance. It involves the following values: +turn taking and f. G the broader social and status roles of the interlocutors. Overall, I considered the status/roles of the represented speakers, taking into account the existing power relations between the narrators as participants and other story participants. In general, DS instances represented interactions between participants with asymmetrical power relations as well as interactions between participants with relatively symmetrical power relations (see also Johnstone, 2002: 112). I therefore came up with the distinction ‘+ or � authority’ in order to define the roles of the represented agents. More specifically, I coded as ‘+authority’ the following represented characters: teachers, police officers, priests, parents, uncles and aunts and unknown adults. In contrast,