Evolutionary Robotics and Interdisciplinary Enactivism

CNRS Summer School:Constructivism and Enaction: A new paradigm for Cognitive Science

Ile d’Oleron, France 29.5.-3.6.2006

Marieke RohdeCentre for Computational Neuroscience and Robotics

University of Sussex

Structure

• Myself

• Past research

• Ideas on Enactivism and Interdisciplinarity

• Ideas for Future Research

Background

• Keen interest in radical constructivism during school and college

• Was caned out of me during my BSc in Cognitive Science in

Osnabrueck, Germany (2000-2003)

• BSc at Intelligent Dynamical Systems group at Fraunhofer AIS, St.

Augustin: First contact with Evolutionary Robotics (2003)

• MSc Evolutionary and Adaptive Systems at Sussex, UK (2003-2004)

• DPhil in Computer Science and Artificial Intelligence at Sussex, UK

(2004 - ???)

Interdisciplinary background, main focus on computational modelling

Computers, Robots, Embodiment: The Braitenberg Vehicle

Braitenberg, V.: „Vehicles. Experiments in Synthetic Psychology.“ illustrations by Maciek Albrecht, MIT Press, 1984

Evolutionary Robotics

Past Work

• Investigation of Motor

Synergies in Motor Control

(Rohde & Di Paolo, 2005)

• A Critique of Value System

Architectures. (Rohde & Di

Paolo, 2006)

– Division between (adaptive)

behaviour generation and

(rigid) adaptation mechanism

• Our model: Evolve a light seeking, fitness predicting robot

– Supports a conceptual argument against “meaning generating boxes”

– Discussion of hybrid/“on-the-fence” position

– Discussion of “neural correlates of behaviour”

Past Work

Interdisciplinarity• Computationalism:

• Roughly: Linear Reduction, AI at

intellectual core

• No real need for dialogue

• Enactivism:

• Mutual Constraining of Disciplines

(as described for

neurophenomenology (Varela

1996)

• Understanding: Hermeneutic

• Introduce computational modelling

into the picture

Future Work

• The “trinity of minimalism”

– Evolutionary Robotics

– Perceptual Supplementation

– Phenomenology

• Investigation: Sensorimotor delays and the construction

of the time arrow

• Inspired by empirical findings on a reversion of the

experienced time arrow after adaptation to increased

sensorimotor delay (D. Cunningham et al.)