CHAPTER 2 LITERATURE REVIEW
2.3 Summary
In this chapter, studies on the issue of viewpoint have been reviewed and the
theoretical background concerning speech-gesture coordination and gesture production has been introduced. Regarding previous studies on viewpoint, scholars have provided copious insights on how viewpoints, or concepts that roughly correspond to the notion of viewpoint, are represented through linguistic and gestural channels. However, the ways in which viewpoints are simultaneously represented through these two channels have not been systematically examined. The purpose of this study, therefore, is to see how viewpoints can both be represented in language and gesture, and how gesture can collaborate with language in expressing viewpoints within the descriptions of third-person past events in conversational contexts.
For the theoretical background concerning speech-gesture coordination process and gesture production, three hypotheses—the Free Imagery Hypothesis, the Lexical Semantics Hypothesis, and the Interface Hypothesis have been introduced. Theories that maintain each hypothesis propose different models to explain how and where gestures are generated, and how gestures are coordinated with their accompanying speech in communicating certain information. These hypotheses might help us to explain the collaborative expressions of viewpoints in language and gesture and also the gesture production process involved in expressing viewpoints with their accompanying speech. In particular, the present study will focus on the discussion on the Lexical Semantics Hypothesis and the Interface Hypothesis. The Free Imagery
Hypothesis, due to its irrelevant prediction on the semantics coordination of speech and gesture, will not be included in the general discussion on the processing of speech-accompanying gesture while collaborating with the speech in expressing the viewpoints.
53
DATA AND METHODOLOGY
In this chapter, the conversational data used in this study, the definition of third-person past events and the selection criteria, the framework adopted, and the qualitative study of linguistic-gestural respective representations of certain viewpoints for use as a basis in analyzing the data of the quantitative study will be introduced.
3.1 Data
The data used in this study comes from five of the conversations in the NCCU Corpus of Spoken Chinese. In particular, the data is selected from the sub-corpus of spoken Mandarin, which contains both oral narratives and daily face-to-face conversations. Concerning the daily face-to-face conversations, participants were college students who knew each other, family members or friends and colleagues. The participants were free to choose any topics which they liked to talk about, and were filmed for about an hour. A stretch from each talk of about twenty or twenty-five minutes, where the participants had already become accustomed to being in front of the camera, was then selected for transcription. The gesture data used in this study comes from the analysis of the gestures made in these transcribed conversations, and
was re-analyzed by other two trained coders separately. Where the two coders were not in consensus on the content of a gesture or where a gesture was not clear, the gesture was excluded.
3.2 Third-person past events and the selection criteria
The present study analyzes the representations of viewpoints by clausal units, since a clause is an independent unit and the smallest grammatical one that can express a complete proposition. And, in terms of the definition of the third-person past events identified in the present study, it refers to a clause that conveys a complete proposition about a third-person’s event that has ever happened in the past. Following Thompson and Couper-Kuhlen’s definition (2005), a clause can be defined as a unit of [predicate+ phrases that accompany it]. The predicate, which refers to a verb or a verb complex, functions to denote and specify the actions being carried out in an event.
The predicate is also the essential element for a unit to be identified as a clause. In this study, there are total of 119 clauses and each of them contains a gesture.
The present study focuses on third-person past events, which means that first-person experiences of the current speaker are not studied. First-person oral experience is excluded in this study since it is not possible to justify the identification of certain viewpoints in this case. It is hard to distinguish whether the speaker is
speaking from the perspective of a current storyteller (which is recognized as the observer viewpoint in this study) or from the perspective of characters in the event (thus character viewpoint in this study), since both storyteller and character are the speaker himself in the rendition of the first-person experience. The “present I” as a storyteller and the “past I” as the character are indistinguishable. To avoid ambiguity in analysis, the present study thus focuses on the study of third-person past events.
Since the study focuses on representations of viewpoint in language and gesture, only clauses with speech-accompanying gestures are chosen. A total of 119 clauses, each clause being accompanied with a speech-accompanying gesture, are analyzed in the present study.
The clauses analyzed in this study include different clause types. 102 clauses in the current data are simple clauses, which account for 85.7% of all clauses. Clauses with the syntactic structure of [S+V+O] and [S+V] are the major types of clause. 46 clauses in the current data follow the structure of [S+V+O] such as tā jiù ná nàge ló
lài bá ‘He holds a grill’, and 31 clauses of [S+V] such as tā jiù chūlái a ‘He just came
out.’ Six clauses of [S+O+V] are of the ba-construction, such as tā jiù bǎ nǐ zhèyàng
cóng dìshàng lā qǐlái a ‘He just pulled you up from the ground like this.’ [O+S+V] is
the structure of topicalization, and one clause in the current data follows this structure.In this clause—Yìmó yíyàng de shìqíng tā yǐqián yě zuòguò ‘The same thing, he has
done before’, the topic Yìmó yíyàng de shìqíng ‘the same thing’ is the object of the verb zuòguò ‘has done’ and appears at clause-initial position.
There are five instances of question in the current data. Two questions follow the common [S+V+O] structure, and two follow [S+V]. The remaining case follows [S+O+V]. Copular clauses, where they appear as [S+copular+NP complement], account for four cases. For example, nǐ shì bóshì bān ‘You are a PhD student’, the subject nǐ ‘you’ and NP complement bóshì bān ‘PhD student’ are joined with the use of the copular shì. Serial verb constructions, which mean that there is more than one verb predicate within a single clause, can also be observed in the current data. Ten cases that follow the structure of [S+VP1+VP2] are identified as serial verb constructions. For example, bǎ tā shēn jìnqù nà yàng <L3 giào giào giào L3> ‘(He) stretched it into the wasp’s nest and stirred it’, two verbs shēn jìnqù ‘stretch into’ and
giào giào giào ‘stir’ are involved within a single clause to suggest two consecutive
actions.
While the majority of the clauses in the current data are single clauses, there are 17 clauses are formed with a complement clause embedded within (14.3% of all clauses). For example, jiùshì zài zhǎo wǒmen jiā dàodǐ zài nǎlǐ zhèyàng zi
‘(Firefighters) were trying to find where our house is’, the subordinate clause wǒmen
jiā dàodǐ zài nǎlǐ zhèyàng zi ‘where our house is’ is the object complement of the verb
in the main clause zhǎo ‘find’. Among these 17 clauses with a complement clause embedded within, 11 clauses have a verb of saying in the main clause. These clauses with a quotative verb such as shuō or jiǎng ‘say’ involve different types of reported
speech, including direct or indirect reported speech. For example, hěnduō lǎoshī shuō
éi zhè kěyǐ názǒu ma ‘Many teachers said, “Can we take this?”’, the complement
clause éi zhè kěyǐ názǒu ma ‘can we take this’ is the quoted speech that follows the verb of saying shuō ‘say’.
Since the clausal events in this study are elicited from conversational data, a clause might sometimes stretch across several turns. Turns of the utterances of other co-conversationalists might intervene between the completion of a clausal event.
Clauses that are formed by intervening utterance turns thus are not selected for analysis in this study. In addition, there are also situations where a clause is a joint utterance. This suggests that one part of the clause is formed by one speaker and the other by another. This kind of clause is also not examined in the present study, since we are looking for the representations of viewpoints for a given speaker in talking about a third-person event. Clauses of joint completion might involve the viewpoints of two speakers.
3.3 Framework of the study
To study the linguistic and gestural representations of viewpoints, a framework which identifies and defines the viewpoints that could possibly emerge from clausal events within conversational contexts should be developed. In communicating third-person past events within a face-to-face interaction, speakers not only recount the propositional contents of the events, but might also simultaneously focus on the ongoing conversation. Several studies have urged the necessity of taking interactive contexts into consideration to study speakers’ footings or perspectives, but have not explicitly established a framework for analysis. Koven’s (2002) framework of speaker role inhabitance was the first study that clearly discussed speakers’ attempts to describe the referential contents of personal experiences and to maintain the ongoing conversations, and a framework for analysis was developed. He explicitly distinguished between three speaker’s perspectives including interlocutory, authorial and character role perspectives, and defined these terms according to the speech events they arise from:
(1) There is the larger context of the interview or conversation in which the stories are told. (Here the relevant speaker roles are those of co-conversationalists or interlocutors.)
(2) There is the event of narration itself, in which the speaker takes on the role of storyteller, obtaining an extended turn at talk in which he/she narrates. (Here the relevant speaker roles are those of narrator/author).
(3) There is the narrated speech event, which is presupposed and invoked in
the event of narration. (Here the relevant speaker roles are those of the narratable and performable characters).
Koven’s framework especially recognized interlocutory role inhabitance as the speaker’s attempt to maintain the ongoing conversation, which was not mentioned in earlier studies. With the interlocutory role perspective being identified, we can see how speakers might also speak on behalf of themselves as the current speaker within a conversation while describing third-person past events. The present study thus follows Koven’s concept of interlocutory role inhabitance and adapts it for convenience to the term speaker viewpoint for the purpose of the present study.
Concerning how speakers present the propositional contents of third-person past events, this study follows McNeill’s (1992) notion of viewpoints and the two viewpoints that he recognized. McNeill has defined the notion of viewpoints as the speakers’ feelings of distance in relation to the past events, and he also brought up two viewpoints—observer and character viewpoint. His discussion on viewpoints, albeit gestures are the main focus, has taken care of both linguistic and gestural manifestations. The present study, since it is also aimed at the study of linguistic and gestural viewpoints, therefore follows McNeill’s notion of viewpoints as a basis for the analysis of viewpoints that arise from the descriptions of third-person past events.
Despite the fact that McNeill’s descriptions of observer and character viewpoint are aimed for use in the study of both linguistic and gestural manifestations, his
definitions regarding these two terms are characterized only in regard to the gestural aspects. In other words, since McNeill’s primary focus is on gestures, he has defined the two terms by describing the differences in their gestural performances. He has defined observer and character viewpoint based on two gestural attributes of iconic gesture—how the gesture represents the characters in narratives and the gestural space used:
Observer viewpoint (O-VPT): The iconic gesture appears to display an
event like the narrator keeps some distance from the story, and the speaker’s hands play the part of the character in the narrative as a whole. The gestural space excludes the speaker’s body.Character viewpoint (C-VPT): The iconic gesture seems to re-enact the
roles of the character in the narrative, and we feel that the narrator is inside the story. The speaker’s hands represent the hands of the character, and the speaker’s body the body of the character. The gestural space incorporates the speaker’s body.Since both observer and character viewpoints are defined by purely gestural perspectives, the ways in which the observer viewpoint and character viewpoint can be distinguished by the use of language are not clearly stated. Koven’s illustrations of
authorial role and character role inhabitance, which in their kernel concepts are
similar to that of McNeill’s observer and character viewpoint, can be used to complement McNeill’s definitions of observer and character viewpoint.
To conclude, building on Koven’s notion of interlocutory role perspective and
McNeill’s terms of observer and character viewpoint, the framework of the present study further identifies three viewpoints—that of speaker, observer and character to account for both linguistic and gestural representations. The definitions can be stated as the following:
Speaker viewpoint (S-VPT): To maintain the larger conversational context
in which the description of third-person past events are embedded or interwoven, the speaker constantly reveals his here-and-now relationship with the co-conversationalists by interacting with them, making comments on, or injecting emotions toward the past events.Observer viewpoint (O-VPT): Concerning the description of third-person
past events, the speaker keeps some distance from the past events, describing the clausal event like an observer outside of the event. The speaker refrains himself either from re-enacting the roles of characters or attempting to maintain the ongoing conversation.Character viewpoint (C-VPT): Concerning the description of third-person
past events, the speaker walks into the temporal and spatial frame of the past events, reconstructing the scene of the past events from the perspective of different characters and enacting their thoughts, speech and other deeds.The analyses of linguistic and gestural representations of different viewpoints in this study will follow the definitions developed here.
3.4 Linguistic representations of viewpoints
In this section, the qualitative study of the linguistic representations of each viewpoint as basis for analysis of the distribution of linguistic viewpoints for
quantitative study will be examined. With respect to the nature of the linguistic representation defined in this study, it refers to a range of multi-layered representations which involves not only grammatical constructions such as sentence forms, phrasal expressions, lexical items, prosodic contours and features, but also other paralinguistic devices such as discourse markers and laughter. Based on Koven’s (2002) analysis of the linguistic representations of three speaker role inhabitance that arise from narratives within conversational contexts, the present study has identified how speakers in talking about third-person past events can also make use of the linguistic and paralinguistic devices recognized by Koven to represent speaker, observer, and character viewpoint. The ways in which these linguistic and paralinguistic devices can represent three kinds of viewpoints will be discussed respectively, following the order of linguistic representations of speaker viewpoint in 3.4.1, representations of observer viewpoint in 3.4.2, and finally the representations of character viewpoint in 3.4.3.
3.4.1 Linguistic representations of speaker viewpoint
The use of various linguistic resources to interact with co-conversationalists, make explicit or implicit comments, and interject their emotions into the described past events all suggest that speakers are showing their here-and-now relationship with
other co-conversationalists, and as such these linguistic resources are seen as representations of speaker viewpoint and are introduced in the following sub-sections.
3.4.1.1 Interrogative sentences
In describing third-person past events, speakers might sometimes choose to use an interrogative sentence as a way of description. By using an interrogative sentence, speakers do not necessarily attempt to solicit more information from co-conversationalists. For interactive function, an interrogative sentence can arouse co-conversationalists’ attention or interest, or their identification toward the described past events. No matter what function it serves, the use of an interrogative sentence suggests that speakers are concerned about co-conversationalists’ epistemic knowledge, attitudes or identification toward the event and it reveals their current status as a speaker within an ongoing conversation rather than that of a speaker simply providing a recounting of the referential meaning of an event. In Example (1), the conversational topic is about how a singer once shared a simple method of preventing excrement from splashing out of the stool. The clausal event being examined is an interrogative sentence where F1 asks her co-conversationalist if the singer had once mentioned a method:
(1) F1: …tā búshì shuō tā yòng shéme fāngfǎ
3SG NEG.COP say 3SG use what method F1: ‘Didn’t he say that he has a method?’
F1’s use of the interrogative sentence is a lead-in to the topic that introduces the singer’s method in the later context to arouse the co-conversationalist’s interest or attention. It can be seen as a signal of the speaker’s attempt to take care of the co-conversationalists, and an attempt to interact with them within the ongoing conversation while describing the events.
3.4.1.2 Speculative expressions
In describing third-person past events, speakers might implicitly inject their own thoughts or opinions, which are purely their own assumptions, toward the past event into the description to explain the initiation or outcome of the events. The use of speculative verbs or adverbs achieves this effect and suggests that the speaker is speaking in his current speaker role, rather than speaking like a storyteller who merely depicts the past events. In Example (2), the conversational topic is about how a group of firefighters help remove the wasp’s nest from the corner of speaker F1’s house. The clausal event being examined describes how the firefighters were able to find her house:
(2) F1: …jiù…(0.7) tāmen hǎoxiàng hái ná le yì zhāng zhǐ EMP 3PL appear still take PRF one CL paper
F1: ‘They appeared to hold a piece of paper in their hands…’
The use of the speculative adverb hǎoxiàng ‘seem to’ suggests that F1 is describing the event using her own supposition.
3.4.1.3 Suggestive expressions
Speakers at times also offer their suggestions or provide ideas on the past events described. In using suggestive expressions, speakers share their opinions toward the event with their co-conversationalists, and at the same time try to elicit the opinions or identifications of the other co-conversationalists. This can often be observed by the adjacency pair of a turn which involves a clausal event and its second pair part (SPP) by a co-conversationalist. In Example (3), speaker F1 offers a suggestion about how her friend’s father tried to use some socks as gloves to remove the wasp’s nest.
(3) 1 F1: …ye jíshí yīnggāi yào ná nàge sùjiāo de nàge a PRT actually should want take that plastic DE that PRT 2 F2: …dùi a wǒ yě bù dǒng ye…
right PRT 1SG also NEG know PRT F1: ‘Eh, he should have used plastic gloves.’
F2: ‘Yeah…I don’t know.’
With the suggestive expression yīnggāi ‘should’, F1 is sharing her ideas with F2 and F2’s response dùi a ‘Yeah’ suggests her agreement with F1. The use of suggestive expressions helps speakers to reveal their current role as a speaker in a conversation
and to interact with other co-conversationalists, respectively.
3.4.1.4 Parenthetical remarks
Speakers at times might explicitly state that they are making comments or providing their own annotations on past events by the use of a parenthetical remark.
The use of a parenthetical remark suggests that the speaker is speaking on behalf of himself as a current speaker in the conversation. In Example (4), speaker M1 is talking about how the president of his company had deliberately not invited some people to a feast. With the use of a parenthetical remark lǎoshí jiǎng ‘frankly speaking’, M steps out of the description of the past event and makes a here-and-now annotation:
(4) M1: lǎoshí jiǎng zhēnde méi dàolǐ frank speak really NEG sense M1: ‘Frankly speaking, it didn’t make any sense.’
3.4.1.5 Evaluative and emotive expressions
In addition to the use of a parenthetical remark that explicitly states that they are
In addition to the use of a parenthetical remark that explicitly states that they are