Conversation analysis of online conversations

In Chapter 4, we would look more closely the concept of turns and how they are split into smaller units in speech Chafe’s intonation units, then in IM transmission units.. Also, the aver

Trang 1

communicate online despite this “lack”. Far from being limited or

impoverished, this online medium is rich with possibilities as users continue

to adapt and use the resources creatively. Werry (1996:47) captured the essence of conversations in this relatively new medium aptly:

interactions online.

Trang 2

Situated in the tradition of conversation analysis (CA), this research aims to discover and present patterns of interaction in online conversations

pertaining to the structural aspects of conversations in IM. In particular, we will see how MSN users continue to build on and use organization strategies employed in face‐to‐face interactions and where regular use is disrupted due

to the constraints within the CMC mode, adapt and find new ways to achieve the same end.

The data is obtained from a closed group of nine people who meet in person regularly. They were instructed to auto‐archive their online conversations on the popular IM programme MSN Messenger for this study. Each archived conversation is a transcript, a copy of how talk unfolded for at least one of the participants in the conversation.

The transcripts were read and where the conversation does not flow or make sense, an explanation was sought by appealing to a variety of theories and analytic tools found in textual analysis or conversation analysis. The analyses were done mostly by observation. Features unique to online conversations

Trang 3

In the next chapter, look at what past research has uncovered about

structural features of IM. Data collection and preliminary problems in dealing with the available data will be discussed in Chapter 3. In Chapter 4, we would look more closely the concept of turns and how they are split into smaller units in speech (Chafe’s intonation units), then in IM (transmission units). The organizational functions of transmission units will also be explored. In

Chapter 5, we will explore if the concept of “overlap” is relevant to IM in MSN. If they do, what are some of the resources that users employ to make sense of and repair overlapping talk? Following that, we will also see how other kinds of repair are done online with the resources available as

compared to face‐to‐face conversation.

We will visit, in the next chapter, a brief overview of earlier studies dealing with computer‐mediated communication (CMC) in general then specifically in

IM.

Trang 4

Walther, 2002) and presentation of self (see Walther, 2007) as compared to face‐to‐face interactions.

Another area of research in CMC concentrate on the kind of language used in CMC interactions by a particular group (e.g. American college students) or across groups (see Prinsen, Volman & Terwel (2007) on gender‐related

differences), many of which are context‐specific, taking place in the

workplace (see Isaacs et al., 2002 and Luor et al., 2010) or classroom (see Kinzie, Whitaker & Hofer, 2005).

Few studies have concentrated on the structural aspects of IM specifically. Below are some of them that undergird this paper, dealing with the

organisational aspects of online talk such as turn‐taking and construction of messages or turns.

Trang 5

Every text has its building blocks. A text is made up of smaller chunks or constituents. In grammatical analysis, the concepts of sentence, clause, phrase and word are used to describe these units. This presupposes that there is a way of identifying these units or where they break off in the larger text.

Baron (2010), while investigating whether the structural features of IM is more like that of spoken language or paradigmatic written language, noted that IM exchanges are broken down into small transmission units (with an average transmission length of 5.4 words) by the users themselves. She observed that entire utterances, complete sentences are broken down by users themselves and sent as separate smaller units to other participants of the conversation.

The same phenomenon is observed in the MSN conversations collected in this study. While Baron (2010) was concerned with the grammatical

relationships between the break pairs (the two chunks1 or constituents which are separated by a break), this study attempts to relate the “transmission unit” to concepts already present in CA. Where do these lines break? What

do these boundaries mark? How are they related to the notions of turns and turn transitions units?

1

Trang 6

no overlap” required of face‐to‐face conversations, where gaps of more than

3 seconds are considered socially awkward (McLaughlin, 1984).

Besides the practical reason of speed of exchanges, there seems to be some relationship between punctuation (in written texts), pauses (in spoken

discourse) and the breaks (in IM). In written text, punctuations are to mark boundaries. When these texts are read, places of punctuation are indicated

by their appropriate length of pauses. In the cases of question marks and exclamation marks, they even indicate emotive intensity and function of utterance, signaled by appropriate intonation.

Halliday (2004) noted the absence of punctuation in the transcription of spoken language into the written from for corpus. The same absence is observed in MSN conversations. It will be reasonable to think that these transition units are thus divided for purposes of punctuation. Also, the

average length of 5.4 words per transmission unit in IM (Baron 2010) is comparable to Chafe’s notion of intonation units (IUs) of roughly 5 words in face‐to‐face conversation (Chafe 1987).

By looking at the functions of intonation units, this study hopes to shed light

Trang 7

transmission units through the aid of the speech act theory (Searle, 1975).

2.2. Talk Management

A preliminary look at the archived IM conversations can prove to be

confusing. Garcia and Jacobs (1999) noted that in quasi‐synchronous IM,

“messages do not display an orientation to their sequential placement in the conversation.” This is because the participants do have any access to what the other parties might be doing – constructing a response to previous

utterance, reading and catching up on previous utterances etc. With short transmission units coinciding with possible turn transition units, there is always a possibility that the next speaker self‐selects and makes a response before the current speaker actually finishes his intended turn or utterance.

To complicate matters, participants are usually multi‐tasking when signed on

to the IM programme at their terminals (Anderson, Beard & Walther, 2010),

it might take a while before the next speaker provides the appropriate

response to prior utterances. In the lapse, other participants may interject and move on away from the current topic.

As a result a highly complex and intertwined dialogue may result. Werry’s (1996) study on interactions on the Internet Relay Chat (IRC) reported:

Trang 8

coherence, turn completions and topic achievements to make sense of

intertwined talk. In the later chapters of this thesis, we will see how these overlaps and interruptions look like on MSN exchanges and what resources the speakers have or depend on to resolve such disruptions.

Prior to any analysis, we will take a brief look at the development of instant messaging and some features of the archived MSN conversations collected in Chapter 3.

Trang 9

on The Microsoft Network (MSN). MSN now co‐exists with Yahoo IM, Skype (for IM and LAN calls) and Facebook2 chat as the main means through which users interact online.

MSN is a one‐way system3 (Herring, 2006) mode of CMC. An entire

transmission unit is sent each time the user chooses to do so. Real time construction of a transmission unit is not displayed in the conversation

window. The programme is designed for “quasi‐synchronous” (not

completely synchronous, since the construction of transmission units is not shown as it is being constructed) interactions though asynchronous

Trang 10

two parties ♪ `lexophile and a bittersweet life as viewed by a bittersweet life.

The nickname and sub‐nickname of the other party is displayed at the top of the window. This makes for clear indication when a user toggles between chat windows and maintains separate conversations with different parties

A toolbar is located just above the text box for the user to use predisposed emoticons, nudges or other animated actions (or “winks”, e.g. rolling on the floor laughing), change their font colour or customize the backdrop display of the chat window.

Trang 11

5

6

43

7

Figure 1: A Typical Chat Window

21

Trang 12

3.2 Data Collection

Trang 14

Original

★Starsparkx in s2★ says (11:59 PM):

i want to watch red cliff 2!

a bittersweet life says (12:13 AM):

I never watch Red Cliff 1 leh

★Starsparkx in s2★ says (12:13 AM):

i want to watch red cliff 2!

I never watch Red Cliff 1 leh

★Starsparkx in s2★ says (12:13 AM):

Trang 15

There were several problems that emerged as I worked with the archived interactions. The first was the lack of detailed and necessary information, usually present and available in transcripts for analysis in the face‐to‐face conversations. The archived online conversations were retrieved from

different machines using different versions of MSN. Although the different versions are still compatible with one another, the recorded data varied with regard to the amount of detail that was recorded. For example, the version used with the MacBook had timestamps down to the seconds whereas the versions used with most other personal computers had it down only to the minute.

The archived conversation preserved primarily the textual elements. Most of the emoticons and small animated clips were not captured. In their place, a blank turn is documented. The final textual exchange is captured with no access to the time taken for production.

Working with the constraints, this study focuses on the aspects of

conversation that are not so dependent on these factors such as sequences, structure of exchanges, repair. Perhaps, a useful question to address at the end of this study is: How do available forms of archiving limit the extent to which meaningful conversation analysis can be done for online

conversations?

Trang 16

or a random technical error. This would mean that some of the conversations archived might not be ‘complete’ in the sense that it is not the same as what the participants had intended.

MSN has a feedback system that informs the sender that their messages

“could not be sent”. But even this feedback system fails from time to time. As

a MSN user myself, I have experienced instances where problems arise in interactions due to missing messages which are unaccounted for.

One way to verify this would be to get the archived conversation from all the parties involved to get an idea of what was intended by all the participants. This was not done in this study because an archived conversation is still a documentation of how the talk unfolded for one of the participants. It would still be interesting to see how the participants come to realize that they are not receiving all the messages that the other party is sending, if such an occasion were to arise.

The third problem relates to the amount of context needed to make sense of

a conversation. The data collected for this study is from a closed‐group of youths who already know each other. The online conversation is a mode of communication, an extension of social activity, a spillover from real life

Trang 17

personally, including but not limited to their personal styles, tones, flair, preference of rhetorics (sacarsm, humour, hyperbole etc.). This knowledge aids them in making pragmatic sense of each other’s utterances.

It is unclear at this point how much of their interpersonal relations need to be brought in to make sense of the data from the researcher’s point of view. The participants are available for clarification of data when the need arises. This issue will be kept in the background and will be revisited at the end of the study to see if it is indeed pertinent.

Trang 18

a bittersweet life says (11:12 PM):

the secular songs are our pursuits of the world: friends, love, school, country, lifestyles etc

we give up the song soon after because we realize these things are not satisfying, we're still lost

Trang 19

how was your posting?

or transmission units are created by deliberate action of the participants even though they have the resources to give longer responses. In other words, the participants have the option of presenting their separate short lines in a longer single line instead. Unless the chunking serves a particular function, the breaks are superfluous.

The transmission units are observable structural units in the archived

conversation. Since the creation of these units (pressing of the ENTER key) require deliberate action by the participants, it will be interesting to take a closer look at where the lines break off, why they break off where they break off and consider what these might mean for online conversations.

Taking the transmission units as the basic unit of organization unit in the

Trang 20

4.2 Transmission Units – What do they do?

The placement of the line breaks is not arbitrary. There are structural and functional principles at work here, sometimes simultaneously in the

Though most of data the do not deviate much from these boundaries, there

Trang 21

Halliday (2004) writing about transcribing the spoken language into the written form for corpus, noticed the absence of punctuation:

Trang 22

The function of punctuation, as suggested by Halliday, marks constituent units and provides prosodic information. The instant messages are similar to transcripts in this aspect, where participants’ talk is presented in print.

Punctuations seem to be abandoned in online talk, as seen in the archived conversations but we also see that the transmission unit breaks are marking constituent units at the word, phrasal, clausal and sentential levels.

Also, Baron (2010) noticed that the average length of 5.4 words per

transmission unit in IM is comparable to Chafe’s notion of intonation units (IUs) of roughly 5 words in face‐to‐face conversation (Chafe 1987). Chafe (1987, 1994) argues that prosodic units (intonation units or IUs) are

cognitively motivated, that the observed IUs are shorter than a clause and represents the limitation on the amount of information that speakers and listeners can deal with at one time. Do transmission units represent the same limitation? If talk online is “edited” rather than “spontaneous”, “read” rather than “heard” (Garcia & Jacobs, 1999), prior information can be retrieved by scrolling up chat log and consulting the earlier exchanges. Why are

transmission units generally short, comparable to the length of IUs?

Efficiency of Lines and Information Packaging

As mentioned in Section 4.1, the transmission units are usually short and there are a couple of pragmatic reasons for this:

Trang 23

1 There is already a time lag between thinking and typing, the

programme (based on the principles of word processing) makes editing possible before the messages are finally sent out by the user. Keeping the transmission units short would cut down the time spent

by the participants in waiting for the current speaker to complete his/her response.

2 Short transmission units facilitate fast and easy reading, reducing the time required for the next speaker to make a response. The short lines allow information to be packed in a way that it is digestible and easily accessible by participants in the conversation.

In other words, the short transmission units (like the IUs) could represent the limitation on the amount of information that speakers and listeners can deal with at one time in the text‐based interactional medium.

On Information and Information Status

Prince (1981: 224) wrote on conveying and packaging of information in language:

Trang 24

Brown & Yule (1983:155) presenting the Hallidayan view of information organization in spoken discourse, wrote:

Trang 25

While the Hallidayan model of information structure is based on prosodic features, Prince (1981) developed it on the syntactic resources. She focused

on reference chains (necessarily restricting it to noun phrases) in particular and found correlations between the speaker‐perceived information status (new, inferrable and evoked entities) and the syntactic forms (e.g.

definite/indefinite expressions, pronominals, pro‐verbals) that speakers employ.

Trang 26

eunice was using this noun phrase to encapsulate the content that a

bittersweet life was articulating directly before this turn, which puts this noun

phrase into the category of “inferable entities”.

Trang 27

The next mention of “nice concept” took the pronominal form “it” in the following transmission unit. Having just been mentioned, this noun phrase is current and still active in the minds of the speaker and falls under the

category of “evoked entities”, its referent textual rather than contextual.

media and communication” and “hahah” to complete her response.

(5)

how was your posting?

mabel says (11:03 PM):

oh got into np!(:

Trang 28

Reference Chain 1 – mabel (the person)

Your Pronominal Evoked entity: situational (defined by

context) (I) Pronominal Evoked entity: situational (defined by

context)

The reference chain for the entity mabel (the person) is not of interest here

as the two referring expressions were used by two different participants and necessarily placed in separate transmission units. Instead, we focus on the next reference chain, where the two referring expressions for the same

New entity: brand new

chinese media and

communication

proper noun (definite)

Inferable entity (a specific course in the school)

In response to a bittersweet life’s question “how was your posting?”, mabel

Trang 29

information status, the noun phrase “np” (an abbreviation for the proper

noun Ngee Ann Polytechnic) is perceived by mabel as brand new information for a bittersweet life. The next noun phrase in the reference chain “chinese media and communication”, which also pertains to mabel’s school posting, is

considered an inferable entity, since “np” has just been brought up earlier and the connection between the course and the school is easily established.

Transmission Unit: Function and Action

In the excerpt below, ervine responds to a bittersweet life’s question and

does another transmission unit consecutively, with a single word “myself” following from “scold la”.

Trang 30

wah, why your sub-nick so fierce?

ervine; says (1:09 AM):

ervine; says (1:10 AM):

always scared i scratch my racket

do with talk. ervine’s first transmission unit is a direct response, an answer to

a bittersweet life’s question. His second transmission unit is appended to his

first to clarify (probably as an afterthought) that he was the object and target

of his scolding, the content of which is displayed on his sub‐nickname5 on MSN.

The packaging strategy facilitates the fulfillment of the intended function in a

way that is efficient and sufficient. In this case, ervine’s next unit takes on the

function of clarification. It was necessary and a single word was sufficient to fulfill this function.

Each transmission unit takes on a function intended by the speaker. We also see strong evidence of that in the way laughter and hesitation is presented in

5

Trang 31

come back to it when you're older

haha

♪ `lexophile says (11:45 PM):

i'll take it as they never had authorial intent

or maybe it's just that some ppl are hard-wired that way

Hesitation

In excerpts (8) to (10), all the instances of hesitation start a turn in response

to a question or suggestion by the previous speaker. Whatever follows the hesitation (er, erm, hmm, hmmm) is placed in the next transmission unit.

(8)

when was that post again?

♪ `lexophile says (1:27 AM):

erm

juno 2006

Trang 32

*june

he lived somewhere near me

_

(9)

I think we're going to talk to him

yeah you write something lah

maybe you could make us do langsong

hmmm

maybe, see how lah

Austin is known for his influential work Doing Things with Words. He noticed

Trang 33

utterances ‐ the locutionary, the illocutionary and the perlocutionary force. The locutionary force of an utterance is its literal or surface meaning, the content of what is being said. The illocutionary force refers to the real or intended meaning of the utterance and the perlocutionary force is the effect

Trang 34

We revisit the earlier excerpt but instead of analysing it with an interest in information status, we look at the function, the illocutionary act of each of

awe/admiration nice concept a representative to provide reason for

awe/admiration who came up with it? a directive

(more specifically, an interrogative with a pragmatic focus on WHO)

to elicit information from hearer

In excerpt (12), eunice has just landed in London early winter and a

bittersweet life is asking if eunice has brought enough clothing over to keep

warm.

Trang 35

so did you bring enough clothing?

and my clothes are only good enough for the current temperature

not really actually a representative to respond to question posed

im alr running out of clothes a representative to inform on the situation and my clothes are only good

sigh i need to go to sch tmr and the day after

how come?

extra lessons and council gm

Trang 36

doing twelfth night

sch: school tmr: tomorrow gm: general meeting

_

doing twelfth night. a representative to elaborate on the situation

Elements at work

In the previous excerpts, we have been able to attribute one function to each line. However, it is conceivable that a unit of information can be used to do

for one i think the range of notes that i wrote for tenors was perhaps too

challenging

yup

normal part writing, I keep it to F

Trang 37

that's like MAX

for now

_

normal part writing, I keep it

to F

a representative to assert a state of affairs

that’s like the MAX an representative to indicate truth of assertion

oh could you send me the minus one for jok2k

*joy2k

Trang 38

YUP!

I wanted to let you hear

the completed version

oh you got it!

I wanted to let you hear A representative to inform of intention

Trang 39

Thus far in this chapter, we have explored the transmission unit as the

smallest organization unit of MSN conversation. The ‘redundancy’ of breaking

up talk into the smaller sub‐units led the inquiry into what the transmission unit does for information packaging and later the role of the unit in shaping talk through their illocutionary force.

It would be interesting to see how the transmission units relate to the

concept of turns in conversation. Can each transmission unit be considered a turn? Can we take a transmission unit to be a turn‐construction unit (TCU)?

According to Sacks et al. (1974), a TCU is defined as a segment of talk that can

be taken as a whole turn. It has two key features:

1 Projectability

– it is possible for one to tell how the turn is going to end before it ends

2 Boundaries (transition relevance places TRPs)

– points where another speaker can start a turn or current speaker may continue with another TCU.

Trang 40

you going to ervine's house tomorrow ah?

Try to lift yourself using your shoelaces says (1:03 AM):

thats like more than 500 sing

tsk tsk

eunice says (6:18 PM):

and i its not on shopping

Định dạng
Số trang	96
Dung lượng	776,94 KB