Prior research into embodied interface agents has found that users like the
m and find them engaging. However, results on the effectiveness of these in
terfaces for task completion have been mixed. In this paper, we argue that
embodiment can serve an even stronger function if system designers use actu
al human conversational protocols in the design of the interface. Communica
tive behaviors such as salutations and farewells, conversational turn-takin
g with interruptions, and describing objects using hand gestures are exampl
es of protocols that all native speakers of a language already know how to
perform and can thus be leveraged in an intelligent interface. We discuss h
ow these protocols are integrated into Pea, an embodied, multi-modal interf
ace agent who acts as a real-estate salesperson, and we show why embodiment
is required for their successful implementation. (C) 2001 Elsevier Science
B.V. All rights reserved.