I had wanted to read this for some time because, as Dale pointed out in his blog post on the Bansho Monograph, there seems to be a lot of sitting and focused conversation required in order for rich learning and consolidation of new understanding to take place during a well-structured, three-part, problem-based lesson in mathematics. Although the Grand Conversations monograph deals primarily with literacy text-based dialogue, it does address "sowing the seeds of deeper comprehension", so I figured the concepts could easily be applied to math conversations.
In my classroom, I have found that having students sit on carpet and listen -- even for short periods of time, and with strategies like Think-Pair-Share built in -- often results in a string of disconnected, inarticulate thoughts. It is at best what the monograph refers to as "gentle inquisition", where I the teacher ask a question, one student responds, or several respond, but none really listen to one another, or build on each other's ideas.
I want to change this!
The monograph suggests some class norms (based on Sipe's 2006 work) for group "discussion" as they call it (based on the intended goals, one might argue "dialogue" to be a better label -- see Garmston and Wellman for more on this):
Like the different sections of Boushey and Moser's Daily Five, the facets of rich talk must also be modelled and practised individually before they become automatic in the classroom.
I see I have my September cut out for me!