SFB 673 - Alignment in Communication

Lecture

Addressees Shape Speaking: When Confederates May Be Hazardous to Your Data	Donnerstag, 08.01.2009 14:00 - 16:00
Anna K. Kuhlen, Stony Brook University, NY, USA Language use and processing can be shaped by coordinated action among conversational partners. Recently, mainstream psycholinguistics research has recognized the importance of studying language in interactive dialogue contexts. For a variety of reasons, experiments often use confederates instead of naive conversational partners. Under what circumstances can the role of a conversational partner be fulfilled just as well by a confederate? And under what circumstances might a confederate change the nature of the dialogue itself? The answer depends on what the essence of spoken dialogue is assumed to be. One common concern is that a confederate's status should be covert. We suggest that a more serious issue is that even confederates blind to experimental hypotheses may know too much, behave accordingly, and naive partners may pick up on this, adapting in undesirable ways. We propose that the danger of using confederates varies, depending on task roles and the phenomena studied. Understanding whether and how confederates can fulfill the role of a conversational partner has important methodological and theoretical implications. From a methodological perspective it is important to understand how confederates can be used without altering the behavior under study. From a theoretical perspective it is important to gain a more thorough understanding of the role a conversational partner has in conversation, more specifically, the processes by which conversational partners adapt to each other.

Donnerstag, 08.01.2009 14:00 - 16:00

Anna K. Kuhlen, Stony Brook University, NY, USA

Language use and processing can be shaped by coordinated action among conversational partners. Recently, mainstream psycholinguistics research has recognized the importance of studying language in interactive dialogue contexts. For a variety of reasons, experiments often use confederates instead of naive conversational partners. Under what circumstances can the role of a conversational partner be fulfilled just as well by a confederate? And under what circumstances might a confederate change the nature of the dialogue itself? The answer depends on what the essence of spoken dialogue is assumed to be. One common concern is that a confederate's status should be covert. We suggest that a more serious issue is that even confederates blind to experimental hypotheses may know too much, behave accordingly, and naive partners may pick up on this, adapting in undesirable ways. We propose that the danger of using confederates varies, depending on task roles and the phenomena studied. Understanding whether and how confederates can fulfill the role of a conversational partner has important methodological and theoretical implications. From a methodological perspective it is important to understand how confederates can be used without altering the behavior under study. From a theoretical perspective it is important to gain a more thorough understanding of the role a conversational partner has in conversation, more specifically, the processes by which conversational partners adapt to each other.