conversation analysis

Speculation about CA mental state | Mike McDonald | June 8th, 2003

I am somewhat confused about the methodology of conversation analysis. In Task 4.2 of the IIC module, on p. 4.33, Keith quotes Levinson's analysis of an interaction between someone requesting a meeting and someone else answering that request. Levinson states that a 2-second silence by the latter is interpreted by the former as a prelude to a refusal; in other words, Levinson attempts to look into the requester's state of mind. On the following page, an analysis by Taylor and Cameron criticizes this as "stepping beyond the methodological limits".

Then, in Keith's comments on p.4.35, he refutes Taylor and Cameron's criticism, saying "where participants. . . are orienting themselves to (normative) features . . . we are justified in offering an account of their actions based on this".

However, in his comments on Task 4.3 (on p. 4.38), Keith says that "speculation about A's mental state has no relevance to the analysis". This seems to contradict his previous comments.

In carrying out Task 4.3, I several times attempted to interpret the participants motives, as in the following examples:

Line 3

In response to B's unusual greeting, A remains silent at the next TRP. His silence may be attributable partly to surprise at the departure from the norm and partly to an inability to rapidly formulate a suitable response. In a later turn (lines 7-8) A explains that he has never been called Jimbo; with this foreknowledge, we may guess that A is faced with a dilemma in line 3: should he treat B's address as a case of rudeness, humor, or mistaken identity?

Line 6

B immediately follows up A's query with "Oh, it's you", pretending to interpret the stress on "baby" rather than "Jimbo" as acknowledgment by A that the latter is indeed a name by which he is known. B's own utterances continue to be very informal, with an inappropriate level of familiarity that clearly marks them as humorous.

Lines 7-8

From line 6, A immediately surmises that B's intention is humorous, and that the humour hinges on his own formal tone. He decides to follow this conversational trajectory, and responds with a ludicrously stilted objection to B's mode of address, using strongly marked formal lexis ("appellation", "hitherto", etc.). The slight hesitation and repetition of "I'm not" may be due to the mental processing involved in creating the long and complex utterance that follows. Although he knows that B's intention is humorous, A is still uncertain why B chose to use the name "Jimbo", and his objection, while humorous, is intended to ascertain the reason for this.

Have I "stepped beyond the methodological limits" or not? Keith's own analysis seems much less speculative, but I feel it ignores some salient features of the interaction.

Mike McDonald

Approaches to DA | Jake Kimball | Sunday, June 8th, 2003

Hi Mike,

I'm happy to take a stab this, as IIC is an on-going module of mine. It's given me a good chance to go back and review the material on 4.8-4.12.

As far as overstepping methodological boundaries, it's possible or even probable. In interpreting the data, have you strayed from basic considerations #2 & 3 on page 4.10? Is there enough contextual information available to analyze the text without overstepping methodological limits? I think the more contextual information we have the less likely we are to 'reach out.'

It's my understanding that CA's approach to analysis lies in interaction as it is unfolding. EC emphasizes context more than CA.

In interpreting the data I can't help but violate the basic considerations by relying on scripts--office and radio shows.

My interpretation of the data is a bit different.

In task 4.2, I recall having difficulty as well. The TRP, evidenced by the lapse in time, can be attributed to a dispreferred answer. I remember wanting to see a video to see what else was happening. What would body language tell us that isn't being spoken? Could it be that C is pleading for an appointment? The 2.0 pause made by R could be 'just a minute, let me check my book.' The 'probably not' could just be a sly deference to a superior (who's holding the power?) in order to secure a meeting. If I'm not mistaken this kind of tactic is promoted in Dale Carnegie's classic How to Win Friends and Influence People. In addition, look at the 1.0 pause after Monday (T1), indicating that C is being very careful, respectful, and indeed has a question for R. In this case, the 2.0 lapse could possibly be R's hesitance in light of C's hidden agenda. Maybe R is thinking 'what exactly does C want from me?' R's response 'hmm yes' also doesn't explicitly answer whether or not he/she is busy, but that he's willing to meet--maybe due to C's formal request via politeness and hesitation in T1 & T3.

I think I've also overstepped methodological limits by reading too much into the relationships of the 2 individuals and the roles they take.

Task 4.3 I also came up with a different interpretation, based on (a priori) script. Line 5 I read as a rhetorical question, highlighting B's inappropriate, informal greeting. Line 6 indicates A's ignorance that he's made a faux pas. 'Right' I'm unsure of. Or maybe he knows he's made a social blunder but by 'right' maybe he wants to avoid the confrontation and move on. 'Oh it's you' back tracks to the previous silence in line 3 in that B wasn't sure that he was indeed talking to A and not that Hookvale character or they weren't connected. Line 7 could be interpreted as 'I'm not 'Jimbo' and don't talk to me as if you're Austin Powers, baby.' Instead he caught himself and decided sarcasm would be a higher road to take, hence I'm not--I'm not aware.' Can we record sarcasm on a transcription?

As far as an approach to discourse, task 4.2 seems to be more authentic. Task 4.3 seems to be more of a performance. The unfolding interaction does have other listeners. Because of this kind of context, 4.3 may not allow for 'normative features' to come together.

Jake

Re: CA mental state | Mike McDonald | Friday, June 13th, 2003

Thanks for your thought-provoking response, Jake, and sorry for my long delay in answering. So many other things had to be done. Now it's a Friday evening so, fortified with a stiff G&T, I'll see if I can take the discussion a bit further.

In interpreting the data, have you strayed from basic considerations # 2&3 on page 4.10?

Yes, I think so. But then so does Levinson, seemingly with Keith's approval.

Is there enough contextual information available to analyze the text without overstepping methodological limits? I think the more contextual information we have the less likely we are to 'reach out.'

Don't you find it's sometimes difficult to distinguish between "contextual information" and "speculation"? For example, in his analysis of Task 4.3, line 12, Keith writes: "A's intake of breath indicates that he is about to respond". Is this claim entirely deducible from contextual information, or is it partly speculation? I mean, one could speculate that the intake of breath indicates shock at the revelation of B's secret lover Hookvale, or annoyance at the mention of B's enemy Hookvale, or any number of other things. You make a similar point about several other lines in the interaction, which brings me back to my basic question: to what extent is it permissible to speculate in CA? Clearly, without some degree of speculation it's hardly possible to say anything interesting. But where does one draw the line?

Mike McDonald

DA Process | Jake Kimball | June 17th, 2003

Mike & other interested parties,

Regarding the reduction of speculation in DA...

Once we record and then transcribe, I think meeting with participants to talk about what their state of mind was like at the time, what intentions they had, etc. is essential if we are to avoid speculation. Playing back the recording would help jog their memories. Only at this point, through negotiation with participants, we can safely make a modestly accurate analysis of a conversation. Being directly involved/present in a conversation brings about a different perspective, I think. Even then, there's the matter of the reply, "that's what I said, but not what I meant."

As far as the tasks IIC that we're talking about, it would be wonderful if a recording was included with the module so that we could work on our transcription skills with a variety of ambiguous conversations.

Jake

Re: DA Process | Mike McDonald | June 17th, 2003

Thanks for the advice, Jake. It seems pretty sensible to interview the participants, and in fact that's what I did for the IIC task in the FND module. But is it a normal part of the CA procedure? If my memory serves me right, it's not emphasized in IIC unit 4, and Sacks himself said "I'm trying to develop a sociology where the reader has as much information as the author, and can reproduce the analysis". If the author is going to go back and interview the participants about what they intended, that surely puts him/her in a privileged position vis-à-vis the readers?

Mike McDonald

P.S. I agree with you about including a recording with the module for transcription practice. That's a great idea! How about it, Keith?

Re: approaches to DA | Raymond Sheehan | June 17 th , 2003

Interesting questions, Mike, leading to other interesting questions and answers!

I'm too far removed from the actual details of IIC tasks at the moment to attempt specific answers to your more specific questions. Generally, though, because I'm attempting to incorporate some of the techniques of CA into my dissertation, I'd say that a fundamental question to bear in mind is 'What does CA have in it for me as a teacher-researcher?' I don't see myself as a purist conversation analyst in training here, but rather as a teacher who can pick out some aspects of CA to incorporate them into a wider range that suits my own purposes at a given moment. I find 'thick description' most frequently suitable in that it can marry CA and contextual consideration logically. I find nothing wrong with speculation as long as you underline that you are proposing an 'interpretation' based on rigorous analysis of the linguistic data in your transcription together with the details of the context you might have noted while recording. As long as 'speculation' is not misrepresented as something more factual, when it is really tentative, then it should be all right, shouldn't it?

Raymond

Re: approaches to DA | Mike McDonald | June 17th, 2003

I hadn't really thought about it that way, Raymond, but I suppose that's the most practical approach, and my natural inclination is to agree with you. My only concern about it is that if I start reshaping tools for my own purposes before I can use them the way they're intended to be used, I may end up using a chisel as a screwdriver.

Mike McDonald

Re: DA Process | Jenny de Sonneville | June 17th, 2003

Jake,

At some level we must be honest: interpretation will always be imposing my frame of reference on someone else's words. After having recorded hours of conversations and discussions with teachers in the last month/six weeks for my DISS (exciting, but exhausting!) I find it extremely useful - after having listened once again to the recordings - to check with the teachers before I presume to "interpret" what they mean.

They say something, then I inevitably filter that through my own perception (reference frame, whatever) and interpret it - but my frame of reference (due to my cultural roots, my upbringing, my educational background, the different social, political, economic groupings which I have been part of) will affect my interpretation. Therefore I need to check with them before I presume to put their words into some kind of an "analysis".

Jenny

Re: Approaches to DA | Mike McDonald | June 18th, 2003

Thanks, Keith, for that splendidly comprehensive response. If I might take you up on a couple of points:

1. What counts as psychological speculation in CA? I think this can be answered, though there are grey areas that need to be negotiated. Basically, the distinction is between structural features of interaction and motivations of individual participants. So if all the empirical evidence we have indicates that a sudden intake of breath is treated as "starting a turn" and may therefore be treated as an interruption, we have a structural basis for regarding it as such unless there is evidence that the interactants are treating it as something else.

Can I take it, then, that in CA there are various "default" interpretations of certain phenomena, which can be called upon in the absence of any strong evidence for alternative interpretations? For example, can we draw a standard equation between intake of breath and starting a turn? I don't see any empirical evidence in lines 9-16 of task 4.3 that the intake of breath should be interpreted in this way. If we can plausibly assign to the in-breath either a structural motivation (starting a new turn) or an affective motivation (surprise), there seems to be no a priori reason why the former should be accorded a privileged status. Surely line 12 could be interpreted as a complete turn expressing some emotion; in this case, would the conventions of CA require us to say simply "some emotion", or would any mention of emotions be beyond the pale?

From Colin Graham's email:

09-11 This time B interrupts, in order to draw attention to A's use of "inestimable". This is the first time B has recognised, albeit indirectly, the issue to which A has drawn attention. What happens next is particularly interesting. [Agree so far] B uses this as the basis for a comment about his own location - another famous first may be a reference to Old Trafford, Man U, the cricket ground etc., also in the south. The reference to the 'famous first' then is B being the first person to call A Jimbo, I think, rather than A being first to use "inestimable". Historically, Stockport was always ambivalent about whether it was a separate entity in Cheshire or part of Manchester. The use of deep as a collocation perhaps suggests the idea that speakers in the deep south of the States are occasionally criticized for their (mis)use of appropriate registers, or something about the fact that Davenport would be way south of Stockport, let alone Manchester.

Interesting. I was born in Manchester but don't remember much about it. Like Colin, I assumed that "another famous first" referred to the use of "Jimbo" rather than "inestimable". I must admit, though, that the whole utterance is difficult to interpret. I was surprised to see that neither Keith nor Colin commented on the reasons for the marked stress on "famous" here, or for the stress on both "posh" and "Stockport" in line 14. The whole excerpt contains only four marked stresses, so presumably they have structural significance.

However, my own tentative explanations are probably too speculative to be permissible within the conventions of CA:

"The stress on 'famous' is somewhat difficult to interpret. Its intention seems to be ironic, perhaps playing on a rivalry between different areas of Manchester and mocking the supposed social superiority of the southern suburbs"

". . . By stressing both 'posh' and 'Stockport', however, B implicitly denies that either Davenport or Stockport has low social status."

3. How much detail should be included in the transcript? Keith didn't comment on Jake's suggestion that the module include a tape for transcription practice. I thought it was a good idea, though - what do others think?

5. Should we use CA as it suits us? Like Ray, I think the answer here is yes, so long as you don't claim that this is CA. . . . As far as IIC assignments go, very, very few indeed would count as CA, though many draw on its procedures and descriptive terms.

Thanks for this useful information.

Mike McDonald

Re: Approaches to DA | Mike McDonald | June 19 th , 2003

5. Should we use CA as it suits us? Like Ray, I think the answer here is yes, so long as you don't claim that this is CA.

How about using CA terminology in other contexts? Would it be inappropriate to use terms such as "self-repair", "insertion sequence", and so on in non-CA analyses, or in hybrid analyses?

Mike McDonald

Approaches to DA | Keith Richards | June 19th, 2003

Hi

Thanks to Jenny and Mike for their additional comments. I think Jenny's response is useful because it's a good example of how to use member validation. As I said in my email, although it has no part in CA, it is valuable in other forms of qualitative inquiry.

Mike raises a number of issues (MY RESPONSES IN CAPS):

1. How about using CA terminology in other contexts? Would it be inappropriate to use terms such as "self-repair", "insertion sequence", and so on in non-CA analyses, or in hybrid analyses?

ABSOLUTELY. CA HAS IDENTIFIED CERTAIN STRUCTURAL FEATURES OF TALK THAT HAVE NOW ENTERED THE MAINSTREAM AND I THINK IT'S PERFECTLY LEGITIMATE TO MAKE USE OF THEIR DESCRIPTIVE POTENTIAL.

2. Can I take it, then, that in CA there are various "default" interpretations of certain phenomena, which can be called upon in the absence of any strong evidence for alternative interpretations? For example, can we draw a standard equation between intake of breath and starting a turn? I don't see any empirical evidence in lines 9-16 of task 4.3 that the intake of breath should be interpreted in this way. If we can plausibly assign to the inbreath either a structural motivation (starting a new turn) or an affective motivation (surprise), there seems to be no a priori reason why the former should be accorded a privileged status. Surely line 12 could be interpreted as a complete turn expressing some emotion; in this case, would the conventions of CA require us to say simply "some emotion", or would anymention of emotions be beyond the pale?

I THINK THE SIMPLE ANSWER HERE IS THAT YOU CAN'T DRAW ANY STANDARD EQUATIONS. CA EMPHASISES THAT WHAT MATTERS IS HOW PARTICIPANTS ORIENT TO THE RULES. SO, FOR EXAMPLE, IF A FIRST PAIR PART IS NOT FOLLOWED BY A SECOND PAIR PART (THE NORM) THAT ACTION WILL BE ACCOUNTABLE AND THE ACCOUNTING WILL BE SEEN IN THE INTERACTION. YOU COULD SAY THAT THE PROVISION OF A SECOND-PAIR PART IS A "DEFAULT" OPTION, BUT THIS REPRESENTS IT AS FAR MORE MECHANICAL THAN IS THE CASE. AS FAR AS THE INTAKE OF BREATH GOES, THE SITUATION IS MUCH LESS CLEAR, BUT THE NATURE OF THE INTAKE AND ANY ASSOCIATED NON-VERBAL SIGNALS (NOT AVAILABLE IN THIS CASE) ARE LEGITIMATE ANALYST RESOURCES.

3. [The alternative analysis of the Manchester issue and the presence of stress.]

I FOUND THIS INTERESTING BECAUSE, AGAIN, IT SUGGESTS THAT THERE ARE AREAS (ESPECIALLY WHERE SPECIFIC CONTEXTUAL KNOWLEDGE IS AN ISSUE) WHERE ANALYSTS VENTURE AT THEIR PERIL. THIS IS CLEARLY ONE. THE LINGUISTIC POINT OF THE EXTRACT, I THINK, EMERGES CLEARLY, BUT THIS PARTICULAR EXCHANGE OPENS UP SO MANY ISSUES OF AFFECT, HISTORY (THE PROGRAMME IS EFFECTIVELY ONE IN A SERIES), AND LOCAL CONTEXTUAL KNOWLEDGE THAT IN THE END WE HAVE TO LIVE WITH THE FACT THAT IT REMAINS OPEN TO SLIGHTLY DIFFERENT READINGS. I THINK I'LL GO BACK TO IT AND TRY TO UNPICK THAT.

4. Keith didn't comment on Jake's suggestion that the module include a tape for transcription practice. I thought it was a good idea, though - what do others think?

SORRY, I MISSED THAT (THE SUBCONSCIOUS AT WORK). I DO HAVE A TRANSCRIPTION TASK THAT WORKS WELL WITH GROUPS (WHERE IT'S POSSIBLE TO SHARE THE REALISATION THAT "CORRECT" REALLY ISN'T AN OPTION) BUT MIGHT NOT BE SO GOOD AT A DISTANCE, BUT I COULD TRY IT. I'VE ACTUALLY CONSIDERED AUDIO AND VIDEO CLIPS AS PART OF SOMETHING COMPUTER-BASED (E.G. ACCESSIBLE ON THE WEB OR VIA CD), SO I'LL WORK ON THAT. I'LL TRY TO GET THE TASK UP IN THAT FORM OVER THE SUMMER AND SEE HOW IT GOES.

Thanks again for the contributions.

All the best

Keith

 

Archive Categories