On "uh" and "uhm" and some of the things they are used to do

			 Emanuel A. Schegloff

				 UCLA

My invitation to participate in this occasion remarked that "it would be 
especially interesting and fruitful for us to initiate a dialogue between 
Conversation Analysis and fields more traditionally represented at this 
workshop (e.g., AI, computational linguistics, psychology), to increase our 
understanding of similarities (and perhaps differences) in approach and 
findings," and I have chosen my topic accordingly.  The empirical focus of my 
presentation will be "uh(m);" the "field more traditionally represented at this 
workshop" will be psycholinguistics/cognitive science; among the similarities 
and differences in approach to be taken up will be the study of naturalistic 
data and the use of "corpora," as in "corpus (psycho-)linguistics." The 
theme of the talk (most generally put) is that the natural home of language 
is in talking; that the natural home of talking is in interaction; that talking-in- 
interaction is the product of describable organizations of practice that we 
now know something about; that these organizations of practice engender 
"places" or "positions" in the talk; that virtually everything in conversation 
needs to be understood by reference to both position and composition; and, 
consequently, that a proper understanding of language and of its deployment 
and understanding in the natural world will require coming to terms with the 
practices of talking-in-interaction.