Research
I am primarily interested in cognitive architecture, both as it is found in Nature and as it might be realised artificially. Because I'm committed to the view that cognition and embodiment are initmately related, I also have a strong interest in robotics. Robots, for me, are a vehicle for testing theories of cognition. Of course, if we have a good theory of cognition - one that passes the test of implementability on a robot - then it will help us build better, more intelligent robots. I'm also interested in consciousness, and I see consciousness and cognition as closely related. Finally, in my opinion, to understand these themes properly entails a certain degree of engagement with philosophy, particularly philosophy of mind.
Here are some of the specific themes I'm interested in. (For details consult my publications, especially
Embodiment and the Inner Life (Oxford University Press, 2010).)
- Global workspace theory. This is Bernard Baars's theory of consciousness, which is based on global workspace architecture. The architecture comprises a set of parallel specialist processes and a global workspace. The specialists compete and co-operate for access to the global workspace, which if granted allows them to broadcast information back to the whole set of parallel specialists.
- The simulation hypothesis. This is the idea that the brain is capable of internally simulating interactions with the environment. This capacity can be likened to the operation of the "imagination", and it works by rehearsing trajectories through sensorimotor space.
- Possible neural substrates of a global workspace. In particular, the hypothesis is that the white matter incorporates a limited capacity communications infrastructure that allows information and influence to be disseminated throughout the brain. My interest is in building models of this communications infrastructure using the methods of computational neuroscience.
- Complex networks, in particular small-world networks and modular networks. The brain's structural (white matter) and functional networks are known to exhibit these properties, and they are also related to the generation of dynamical complexity.
- Dynamical complexity. Many systems, including the brain, exhibit very complex behaviour, and it has been proposed that this is the result of a balance of segregated and integrated activity. I am interested in building models of such systems, especially based on complex network topologies and using spiking neurons.
- Robotics. Ultimately, the acid test for all of the above ideas is robot implementation. I am especialy interested in humanoid robotics.
- The frame problem. Originally this was a problem that arose in logic-based AI (which I worked on extensively), but my recent interest relates more to the philosophers' interpretation of the term.
In the late 1980s and 1990s I worked on logic-based knowledge representation, specifically on reasoning about action and the frame problem. I was closely involved in developing and extending a formalism for reasoning about action called the event calculus. A prominent strand of my work concerned the use of abduction with the event calculus, for planning and explanation.