Theory Of Mind For A Humanoid Robot PDF Print E-mail
Written by Rizki Noor Hidayat Wijayaź   

Brian Scassellati MIT Artificial Intelligence Lab 545 Technology Square - Room 938 Cambridge, MA 02139 USA


If we are to build human-like robots that can interact naturally with people, our robots must know not only about the properties of objects but also the properties of animate agents in the world. One of the fundamental social skills for humans is the attribution of beliefs, goals, and desires to other people. This set of skills has often been called a *theory of mind*.

This paper presents the theories of Leslie [27] and Baron-Cohen [2] on the development of theory of mind in human children and discusses the potential application of both of these theories to building robots with similar capabilities. Initial implementation details and basic skills (such as finding faces and eyes and distinguishing animate from inanimate stimuli) are introduced. I further speculate on the usefulness of a robotic implementation in evaluating and comparing these two models.

Introduction Human social dynamics rely upon the ability to correctly attribute beliefs, goals, and percepts to other people. This set of metarepresentational abilities, which have been collectively called a "theory of mind" or the ability to "mentalize", allows us to understand the actions and expressions of others within an intentional or goal-directed framework (what Dennett [15] has called the intentional stance).

The recognition that other individuals have knowledge, perceptions, and intentions that differ from our own is a critical step in a child*s development and is believed to be instrumental in self-recognition, in providing a perceptual grounding during language learning, and possibly in the development of imaginative and creative play [9].

These abilities are also central to what defines human interactions. Normal social interactions depend upon the recognition of other points of view, the understanding of other mental states, and the recognition of complex non-verbal signals of attention and emotional state. Research from many different disciplines have focused on theory of mind. Students of philosophy have been interested in the understanding of other minds and the representation of knowledge in others. Most recently, Dennett [15] has focused on how organisms naturally adopt an "intentional stance" and interpret the behaviors of others as if they possess goals, intents, and beliefs. Ethologists have also focused on the issues of theory of mind.

Studies of the social skills present in primates and other animals have revolved around the extent to which other species are able to interpret the behavior of conspecifics and influence that behavior through deception (e.g. Premack [33], Povinelli and Preuss [32], and Cheney and Seyfarth [12]). Research on the development of social skills in children have focused on characterizing the developmental progression of social abilities (e.g. Fodor [17], Wimmer and Perner [37], and Frith and Frith [18]) and on how these skills result in conceptual changes and the representational capacities of infants (e.g. Carey [10], and Gelman [19]). Furthermore, research on pervasive developmental disorders such as autism have focused on the selective impairment of these social skills (e.g. Perner and Lang [31], Karmiloff-Smith et. al. [24], and Mundy and Sigman [29]). Researchers studying the development of social skills in normal children, the presence of social skills in primates and other vertebrates, and certain pervasive developmental disorders have all focused on attempting to decompose the idea of a central "theory of mind" into sets of precursor skills and developmental modules.

In this paper, I will reviewtwo of the most popular and influentialmodels which attempt to link together multi-disciplinary research into a coherent developmental explanation, one from Baron-Cohen [2] and one from Leslie [27]. Section 4 will discuss the implications of these models to the construction of humanoid robots that engage in natural human social dynamics and highlight some of the issues involved in implementing the structures that these models propose. Finally, Section 5 will describe some of the precursor components that have already been implemented by the author on a humanoid robot at the MIT Artificial Intelligence lab.

[download: 6,15 MB]