You are here

"Goal-directed decision making" explained in a special issue of PhilosophyTransaction of the Royal Society B

1 Oct 2014 - 10:01

Paul Verschure and his collaborators contributed to the edition of the current issue of PhilosophyTransaction of the Royal Society B [established in 1665 by the Royal Society is the oldest scientific journal] with an introduction and the cover of the Journal which was produced by SPECS (graphic contribution Sytse Wierenga).

see  introduction at The principles of goal-directed decision- making: from neural mechanisms to computation and robotics.  Giovanni Pezzulo, Paul F. M. J. Verschure, Christian Balkenius and Cyriel M. A. Pennartz.

The current issue of Phil Trans B brings together scientists from the most relevant research fields in the study of decision making and goal-directed behavior and reports on results and theoretical insights that reinforce its neuronal mechanisms in the brain of humans, monkeys and rodents, what are the computational principles, and its validity in real-world robotic settings.

I do what I do: why and how?

In the same issue, Paul Verschure also authored the article The why, what, where, when and how of goal-directed choice: neuronal and computational principles. Paul F. M. J. Verschure, Cyriel M. A. Pennartz and Giovanni Pezzulo

Research pursued by ICREA prof. Paul Verschure and his group SPECS at UPF has brought new pivotal insights into the understanding of the neuronal mechanisms underlying goal directed behavior and decision-making. In a research paper published in the oldest scientific journal Philosophical Transactions of the Royal Society B 369: 20130470, Verschure and his collaborators including Cyriel M. A. Pennartz University of Amsterdam and Giovanni Pezzulo from the Inst. Of Cognitive Science and Tech., NRC, Rome, Italy, describe how the central problems that goal-directed animals must solve can be summarized as the H4W problem: ‘What do I need and Why, Where and When can this be obtained, and How do I get it?’  Using a neuro-robotic approach Verschure first maps the “brain data” onto the Distributed Adaptive Control embodied cognitive architecture [1] that considers the generation of adaptive action in the real world as the primary task of the brain rather than optimally solving abstract problems. In this paper Verschure and co-authors map the DAC theory towards the different structures in the rodent brain involved in goal-oriented action. The DAC theory emphasizes that perception, emotion, cognition and action are realized through an integrated control system that is both embodied and situated in the real world. From the DAC perspective, drives and goals are necessarily organized in a hierarchical fashion, starting with the concrete needs defined by the physical instantiation of the agent, i.e. the nutrients required to maintain the body, to the abstract goals of reaching specific but also abstract goal states.
The authors analysis of goal-oriented choice, combining a neuro- biological perspective with the DAC theory, exemplifies how neuro-robotic and experimental methods can work hand-in-hand. As the biologically grounded model constitutes an integrated architectural solution to problems of goal- directed choice and pursuit, it can help studying the systems-level neurobiology of goal-directedness rather than focusing only on its components in isolation.