Could next generation androids get emotionally close? `Relational closeness' from human dyadic interactions

Studies of human-human interactions indicate that the relational dimensions encoded nonverbally between people include intimacy/involvement, status/control, and emotional valence. In assessing nonverbal behavior a key issue concerns the correct level or unit of behavior to code. Low-level codes, such as head nods, eyebrow flashes, and smiles, are concrete enough to be specified objectively. However, a coding scheme based on them may not match the phenomenology of lay people's experiences of natural interactions. A high-level code, such as values intimacy, reliably distinguishes secure and insecure attachment styles but is underspecified at the concrete, bodily level. This paper considers what level of behavior codes may realistically be mapped onto next generation androids. New 'mid-level' behavior codes are offered based on an experimental study of relational closeness in human dyadic interactions. These provide act specifications for a possible benchmark of relational closeness. The appropriateness of certain relational behaviors by androids is considered.

[1]  P. Costa,et al.  Validation of the five-factor model of personality across instruments and observers. , 1987, Journal of personality and social psychology.

[2]  R. Heslin,et al.  Meaning of touch: The case of touch from a stranger or same sex person , 1983 .

[3]  Sebastian Thrun,et al.  Toward a Framework for Human-Robot Interaction , 2004, Hum. Comput. Interact..

[4]  Batya Friedman,et al.  Robotic pets in the lives of preschool children , 2004, CHI EA '04.

[5]  D. Berry,et al.  Personality, Nonverbal Behavior, and Interaction Quality in Female Dyads , 2000 .

[6]  R. Sternberg,et al.  People's conceptions of intelligence. , 1981 .

[7]  Hiroshi Ishiguro,et al.  Toward social mechanisms of android science , 2006 .

[8]  T. Kimball,et al.  Observing attachment behavior in couples: the Adult Attachment Behavior Q-Set (AABQ). , 2004, Family process.

[9]  Cory D. Kidd,et al.  An Interactive Robot in a Nursing Home: Preliminary Remarks , 2005 .

[10]  M. Buber,et al.  I and Thou , 1923 .

[11]  Billy Lee,et al.  Interpersonal Perception in Japanese and British Observers , 2004, Perception.

[12]  Ishiguro Hiroshi,et al.  The Study of Interaction through the Development of Androids , 2004 .

[13]  Frank J. Bernieri,et al.  Interpersonal coordination: Behavior matching and interactional synchrony. , 1991 .

[14]  D. Funder,et al.  The Riverside Behavioral Q-sort: a tool for the description of social behavior. , 2000, Journal of Personality.

[15]  Cynthia Breazeal,et al.  Toward sociable robots , 2003, Robotics Auton. Syst..

[16]  Hiroshi Ishiguro,et al.  Toward social mechanisms of android science: A CogSci 2005 Workshop: 25 and 26 July 2005, Stresa, Italy , 2006 .

[17]  P. Costa,et al.  The five-factor theory of personality. , 2008 .

[18]  R. Gifford.  Mapping Nonverbal Behavior on the Interpersonal Circle , 1991 .

[19]  Laura K. Guerrero,et al.  Attachment‐style differences in intimacy and involvement: A test of the four‐category model , 1996 .

[20]  Ross A. Thompson.  Empathy and its origins in early development. , 1998 .

[21]  Karl F. MacDorman,et al.  Androids as an Experimental Apparatus: Why Is There an Uncanny Valley and Can We Exploit It? , 2005 .

[22]  Stein Bråten.  Intersubjective communication and emotion in early ontogeny , 1998 .