10.13: Doing Discourse analysis= 4 approaches

We talked through the 4 sample essays, each of which used a different approach to analyzing the language/information in the transcript.

Smith = social linguistic (interactional) approach
As the name for this approach suggests - the focus was on LANGUAGE, in particular on interactional units.  Smith classified 3 levels of interactions:

• message units = an utterance - something said
• interactional units =  sets of messages that are connected - such as questions and answers
• sets of interactional units = larger clusters of interactions focused on a common purpose, line of exploration etc.

Organizing the trancript in terms of this classification allowed Smith to look at how sets of interactional units interact to create learning situations.

Goldman = cognitive
Goldman looked at the transcript in terms of a model of how individuals use mental represntations to understand the world around them.
The model she used assumed that as we interpret the world, we resort to a model made up of

• a surface layer (the physcial presentation of whatever we are interpreting)
• a textbase = what the text says - as conveyed through the language of the text
• a situation model = or the assumptions referred to by the text - and acted on by the reader= how the reader interprets the text.

Goldman gave evidence from the trancript for the conversations location in a particular level. Her analysis looked at relationships among these different cognitive levels of :"understanding" a text - and noted how the transcript moved sequentially from one level to the next in a way that created effective learning.

Macbeth - ethnometholdological
Macbeth used a close examination of language to establish the values in the conversation.  In particular he looked at how the teacher  signalled "correct" (valued) answers to students, and described the interactions that allowed students to "read" these answers.

Carter - Black feminist, micro-ethnographic
Carter looked at how larger cultural stories (assumptions/values/beliefs => stereotypes) influenced the classroom conversation and what it means that they do. In particular, she noted instances where students gave answers that were NOT supported in the text.  The assumption in this method is that if individuals give answers not warranted by the conversation/text under discussion => that they come from some other cultural source (that validates/suggests them as possible answers).

Sample application of Sociolinguistic/interactive approach:
After reviewing the four approaches, we then began to APPLY the sociolinguistic/interactive approach to the sample transcript on Gaming (posted to the right).

To do this - we used the same analytic approach (classifying the data in the transcript in terms of messages, interactions, and interaction sets; identifying the characteristics of boundaries between the interaction sets; and paying attention the patterns in the transcript that were made visible through marking it up in terms of the sociolinguistic model from the essay.

We got about two thirds through the transcript in terms of identifying sets of interactions - and defining the features that identified boundaries between different interactions.


Blog 9, Part 1:  Finish analyzing the Practice Transcript in terms of  the sociolinguistic approach from Smith's essay.  Your post should include:

• a "marked up" transcript represented in terms of the sociolinguistic approach 
• a list & discussion of patterns that the approach allowed you to see
• a list & discussion of the questions you might answer through extending your application of this method to the transcript.

Example of what I mean by a discussion patterns in the conversation that become visible when the transcript is viewed through this model  = we noticed that Chandler mostly directed the focus of the discussion - but that she also, in some ways, followed the "lead" set up by Brian => conversation was interactive in a way so that Ch both did and did not determine the direction of conversation.  We didn't finish our analysis - so a closer look  may lead to much more significant observations about who directs conversation & how they do it; who "learns" what; how power dynamics play out, and so on).  These observations will set up your discussion of which questions this method could answer effectively.

Blog 9, Part 2: develop another analysis of the Gaming Transcript using one of the three other models.


 On Monday - we will work on these analyses as a class.  The idea is for you to get as much experience as possible trying out these methods.  In addition to posting your analyses on your blog, send me a copy at the course email as a MSword document .  This way I will be able to send feedback & comments.


Format each analysis so that it includes:


Name of the approach you choose

• a "marked up" transcript represented in terms of the model used for the approach
• a list & discussion of patterns that the approach allowed you to see
• a list & discussion of the questions you might answer through extending your application of this method to the transcript.

I will be reading Blogs beginning on Saturday.  Hopefully you can yous this feedback on some of your early efforts at discourse analysis to set you up for the work we will do on Monday.

Good class - and have a great weekend.