Joel Parthemore MA. (University of Sussex)


Joel Parthemore and Ron Chrisley
PAICS Research Group
Centre for Research in Cognitive Science
Department of Informatics
University of Sussex, Falmer, Brighton, UK

We use the term synthetic phenomenology to refer to methods of specifying the contents of experience that go beyond the capacity of language to express. If I want to convey to you the visual experience of a certain agent, I might take a photo or a series of photos; but, to be useful, the pictures would need to be altered in ways that do justice to the manner in which the agent isembodied and embedded in its environment. In the human case, not only does the visual field contain blind spots that are not part of the agent's visual experience; the agent's experience may also contain information that is not currently within its field of view.

This paper addresses the practical considerations of translating a theory like synthetic phenomenology into a working model: in this case, one implemented on an AIBO robot. How are general principles translated into lines of code; how does abstract code play out in real-time robotics; and in what ways does practice have consequences for theory?

By using an artifact that we can interact with and manipulate in a way that we cannot for practical or ethical reasons with living organisms, and without any assumption of the artifact having experience, we can use it to model experience. We can then use the results to feed back in to refining the model.

Among the issues faced in reconstructing the AIBO's "visual experience": the AIBO's "eye" is a single fixed camera located at the midpoint between its two apparent eyes. A human eye saccades without any need to move the head. But the only way the AIBO can perform an equivalent saccade is by moving its entire head. Likewise it is impossible to specify precisely where the AIBO should turn its head, or to determine at precisely what angle its currently oriented.

The presentation will include lessons from personal experience, coding triumphs and pitfalls, and both visual stills and animated montages representing the visual experience that an agent embodied in a similar manner to the AIBO might be said to have.

Joel Parthemore is a second-year DPhil student studying knowledge representation formalisms and the nature of conceptual knowledge at the University of Sussex in Brighton, UK. His research interests include the relationship between representation, resemblance and similarity, as well as the relationship between philosophical theories and implemented computer-based models of those theories. He is interested in tools for helping people constructing external models of their conceptual domains. He is a member of the Philosophy of AI and Cognitive Science research group in the Department of Informatics, and the organizer of the department's E-Intentionality seminar series. Before starting his present course, he taught computer literacy with US Peace Corps in Ghana and worked as an IT specialist with Peace Corps in Washington, DC. In his spare time he plays with Linux computer systems or goes cycling. You can find him online at or