The Kontatti corpus is made of semi-spontaneous dialogic data that allow the research team to work on discourse organization features.
A good method to gather dialogic data from a variety of speakers is that of Map Task. This helps speakers overcome the embarassment of speaking freely in front of a microphone. Besides, Map Tasks provide comparable data as well as fluent dialogic speech. Map Tasks consist in an instruction giving task based on a map: basically, speaker A instructs speaker B to follow an itinerary on a map. Then roles are exchanged. Since speakers don’t see each other’s maps, lots of discussion and negotiation come about, which is good!
We also try to record more freely occurring speech, by letting speakers interact with each other and with the researcher and we sometimes organize data recording parties: dinners, aperitivo, playing cards…
We also require participants to fill in a short sociolinguistic questionnaire in order to sketch a sociolinguistic profile of speakers and speech communities.
You can download our Map Task model here and use it for your project: just let us know how it worked!
You can download here our sociolinguistic questionnaire in Italian, German or Gardenese Ladin: